Docstoc

Made in Madagascar - The Impact

Document Sample
Made in Madagascar - The Impact Powered By Docstoc
					                   MADE IN MADAGASCAR

 THE IMPACT OF RULES OF ORIGIN ON THE TEXTILE AND

                        CLOTHING INDUSTRY


                                     by

                               Anna Andersson




                NATIONALEKONOMISKA INSTITUTIONEN
                      VID LUNDS UNIVERSITET


              Department of Economics at the University of Lund

                                   2009:7

                           Minor Field Study Series

                                   No. 193
Mailing address:                                      ISSN 0283-1589
Nationalekonomiska Institutionen
Box 7082
S-220 07 LUND
Sweden
                                                             Master’s Thesis
                                                                  June 2009




                    MADE IN MADAGASCAR
   The Impact of Rules of Origin on the Textile and Clothing Industry




                                 Anna Andersson




Lund University
School of Economics and Management
Department of Economics                             Supervisor: Yves Bourdet
Acknowledgements

This study was made possible thanks to a Minor Field Study scholarship financed by the
Swedish International Development Cooperation Agency (Sida). I am grateful to Sida for giving
me the opportunity to go to Madagascar and to my supervisor Yves Bourdet for encouraging me
to apply for the scholarship in the first place. I would furthermore like to thank Yves for his
enthusiasm that kept my spirits high even when Madagascar was on the verge of civil war. My
deepest gratitude also goes to Sandrine Rakotovao for getting me the right connections and to all
the other people in Madagascar who took the time to meet me. Thank you as well to Ludovic
Angot for helping me with my French.




                                                                                               2
Abstract

This study investigates the impact of rules of origin (ROO) on the Malagasy textile and clothing
industry. The ROO of two different preferential trading arrangements for developing countries,
the African Growth and Opportunity Act (AGOA) of the US and the Lomé/Cotonou agreement
of the EU, are compared and related to Malagasy clothing exports and textile imports. The
AGOA ROO are found to be more liberal than the ones in Lomé/Cotonou, especially when it
comes to input sourcing. This study shows that strict ROO have a negative impact on Malagasy
clothing exports. Clothing exports to the EU tend to grow slower, be less diversified and use less
diversified inputs than exports to the US. Further, strict ROO are not found to increase vertical
integration in the Malagasy textile and clothing industry. It can therefore be questioned whether
using strict ROO as a tool for development policy in highly fragmentized sectors is effective.
Lastly, the utilization rates for AGOA apparel are higher than the ones for Lomé/Cotonou
apparel, reflecting the higher costs of exporting to the EU due to strict ROO. In conclusion, the
limited input sourcing possibilities in the Lomé/Cotonou ROO can be said to have limited the
expansion and diversification possibilities of the Malagasy textile and clothing industry.




Keywords:
Rules of origin, Madagascar, textile & clothing, AGOA, Lomé/Cotonou




                                                                                                3
Table of Contents
    List of Figures ............................................................................................................................. 5
    List of Tables............................................................................................................................... 5
    Abbreviations .............................................................................................................................. 6

1 INTRODUCTION .................................................................................................................... 7
  1.1 Purpose .................................................................................................................................. 8
  1.2 Delimitations ......................................................................................................................... 8
  1.3 Method and Material ............................................................................................................. 9
  1.4 Disposition ............................................................................................................................ 9
  1.5 Previous Research ............................................................................................................... 10

2 MADAGASCAR AND TEXTILES...................................................................................... 12
  2.1 Madagascar’s Economic Background ................................................................................. 12
  2.2 The Textile and Clothing Sector ......................................................................................... 13
     2.2.1 Export Processing Zone ............................................................................................... 15
     2.2.2 The Multi Fiber Agreement.......................................................................................... 17

3 THE ECONOMICS OF RULES OF ORIGIN .................................................................... 20
  3.1 Background to Rules of Origin ........................................................................................... 20
  3.2 Rules of Origin - a protectionist tool?................................................................................. 23
     3.2.1 Rules of Origin as transaction costs ............................................................................. 23
     3.2.2 Increased production costs through supply-switching ................................................. 24

4     PREFERENTIAL TRADING ARRANGEMENTS ......................................................... 29
    4.1 EU Preferential Trading Arrangements............................................................................... 29
       4.1.1 Lomé and Cotonou ....................................................................................................... 29
       4.1.2 Rules of Origin for Textiles and Clothing.................................................................... 30
    4.2 US Preferential Trading Arrangement ................................................................................ 31
       4.2.1 African Growth and Opportunity Act (AGOA) ........................................................... 31
       4.2.2 Rules of Origin for Textiles and Clothing.................................................................... 32

5    IMPACT OF AGOA AND LOMÉ/COTONOU ................................................................ 34
    5.1 Response to AGOA............................................................................................................. 34
    5.2 Exports to the EU ................................................................................................................ 38
    5.3 Textile Inputs....................................................................................................................... 41
       5.3.1 Origin of Imported Inputs ............................................................................................ 42
       5.3.2 Domestic Inputs............................................................................................................ 47
    5.4 Utilization Rates.................................................................................................................. 48

6    CONCLUSIONS.................................................................................................................... 50

REFERENCES ............................................................................................................................ 53

APPENDIX .................................................................................................................................. 59

                                                                                                                                                 4
List of Figures
Figure 2.1 - Madagascar’s annual GDP growth in % 1990-2006 ................................................. 13
Figure 2.2 - Textile and clothing exports as a share of total export .............................................. 14
Figure 3.1 - Value content ROO and Costs................................................................................... 25
Figure 3.2 - Hub-and-spoke system .............................................................................................. 27
Figure 5.1 - Clothing exports to the EU and the US ..................................................................... 34
Figure 5.2 - Exports of Clothing to the US ................................................................................... 36
Figure 5.3 - Exports of Clothing to the EU ................................................................................... 39
Figure 5.4 - Total import of textile yarn and fabrics 1990-2006................................................... 42
Figure 5.5 - Total import of textile yarn and fabrics 1990-2006 by origin ................................... 43
Figure 5.6 - Textile yarn and fabric imports from Africa 1990-2006 ........................................... 43
Figure 5.7 - Textile yarn and fabric imports from the EU 1990-2006 .......................................... 44
Figure 5.8 - Textile yarn and fabric imports from Asia 1990-2006 .............................................. 44
Figure A.1 - Composition of Malagasy imports of textile yarn 1990-2006.................................. 60




List of Tables
Table 2.1 - Top 5 exported products in 2006 ................................................................................ 14
Table 5.1 - Most exported clothing categories to the US in USD million .................................... 37
Table 5.2 - Most exported clothing categories to the EU in USD million.................................... 40
Table 5.3 - Utilization rates for AGOA and Lomé/Cotonou apparel 2002-2007 ......................... 49
Table A.1 - Clothing exports of African and Asian producers in million USD after the MFA phase-out ...... 59




                                                                                                                                 5
Abbreviations
ACP           African, Caribbean and Pacific
AGOA          African Growth and Opportunity Act
ATC           Agreement on Textiles and Clothing
ASEAN         Association of South East Asian Nations
CACM          Central American Common Market
ESA           Eastern and Southern Africa
EPA           Economic Partnership Agreement
EEC           European Economic Community
EU            European Union
EBA           Everything But Arms
EPZ           Export Processing Zone
FTA           Free Trade Agreement
GATT          General Agreement on Tariffs and Trade
GDP           Gross Domestic Product
IEPA          Interim Economic Partnership Agreement
IMF           International Monetary Fund
LDC           Least Developed Country
LTA           Long Term Agreement Regarding International Trade in Cotton Textiles
MFN           Most Favoured Nation
MFA           Multi Fiber Agreement
OTEXA         Office of Textiles and Apparel
PTA           Preferential Trade Agreement
ROW           Rest of the World
ROO           Rules of Origin
SADC          Southern African Development Community
SITC          Standard International Trade Classification
T&C           Textile and Clothing
UN COMTRADE   United Nations Commodity Trade Data Base
UNCTAD        United Nations Conference on Trade and Development
US            United States
USD           United States Dollar
VAT           Value Added Tax
WTO           World Trade Organization




                                                                                     6
1       Introduction


Rules of origin (ROO), which are an essential tool in international trade, determine the
geographic origin of goods. The policy usages of ROO are diverse and include for example
collecting trade statistics, applying import tariffs and safeguard measures, marking products as a
service to consumers and imposing anti-dumping duties on unfairly traded goods. In addition,
ROO are also used in Preferential Trade Agreements (PTAs) to ensure that only the intended
countries benefit from preferential treatment.1 Globalization has lately made production
processes more fragmentized and increased the number of PTAs in the word rapidly.2 This has
made origin marking harder and in turn highlighted the importance of ROO.


The increasing importance of ROO has meant that more and more attention has been diverted
towards them. Their supply-switching effects through inflexible input sourcing rules and
administrative costs have been found to have a negative effect on trade.3 Questions about how
trade enhancing a PTA with strict ROO really is have therefore been raised. Rules of origin also
reduce the value of the tariff concessions in preferential trading arrangements for developing
countries.4 This is problematic since it limits trade creation and development for poor countries
and hence counteracts the whole purpose of these trading arrangements. Critical voices have
even accused ROO of being a form of hidden protectionism.5 These accusations make it
interesting to examine to what extent ROO can affect trade between countries in general and
especially in the case of preferential trading arrangements for developing countries.


Madagascar is one example of a developing country that is dependent on preferential trading
arrangements. In order to reduce poverty, Madagascar has chosen an outward-looking
development strategy where the Export Processing Zone (EPZ), Zone Franche, has played a
crucial role. The Malagasy EPZ is one of few economic success stories in sub-Saharan Africa


1
  Ahmad (2007) p. 1
2
  Ghoneim (2003), p. 598
3
  Augier, Gasiorek and Lai-Tong (2005) p. 576
4
  Brenton and Manchin (2003) p. 767
5
  Ahmad (2007) p. 1
                                                                                                7
and the reason behind the success is foremost a booming Textile and Clothing (T&C) sector. The
sector’s expansion has been heavily reliant on the preferential access to both the European Union
(EU) and the United States (US) granted to Madagascar for being a Least Developed Country
(LDC). The ROO for clothing in the EU and the US preferential trading arrangements differ
however. The EU rules have traditionally been quite strict while the US rules are known to be
liberal, especially in terms of input sourcing. This fact makes Madagascar an excellent candidate
for a case study on the impact of ROO on the T&C sector.



1.1 Purpose
The purpose of this study is to examine the impact of ROO on the T&C sector in Madagascar.
The analysis focuses on final good exports and input imports to study the effects of liberal versus
strict ROO and aims at answering the following:
   •   Is it possible to see different trade patterns with the US and the EU when it comes to
       Malagasy clothing exports? If so, what are the differences and can they be related to
       ROO?
   •   Further, can ROO be deemed to influence the choice of inputs in the Malagasy T&C
       industry? In what way?
   •   To what extent is preferential access requested for Malagasy clothing exports to the US
       and to the EU? Is there a difference between the utilization rates for AGOA and
       Lomé/Cotonou apparel?



1.2 Delimitations
The paper is mainly based on data up until the year 2006. This means that the effects of this
year’s political crisis in Madagascar and of the new Interim Economic Partnership Agreement
(IEPA) between the EU and Madagascar are not covered by the data and hence not analyzed
further. The IEPA is however mentioned in chapter four and chapter six includes a short update
to recent events and their possible long-term effects for Madagascar.




                                                                                                 8
1.3 Method and Material
The study is based on economic theory in relation to economic integration and ROO. A
comparative method is used to examine the differences, if any, between the trade effects of the
US and EU ROO. The trade effects are analyzed by looking at export and import responses to
the different preferential trading arrangements. In particular, the composition of final good
exports to different markets is examined and related to the imports of input material. The data is
collected from the United Nations Commodity Trade Data Base (UN COMTRADE), the EU’s
Eurostat and the US Office of Textiles and Apparel (OTEXA). The classification system used for
data collection from UN COMTRADE is the Standard International Trade Classification (SITC)
revision 2. This older system is used in order to get long time series on a reasonably
disaggregated level. Utilization rates of the preferential trading arrangements and an examination
of domestic input alternatives also serve to deepen the understanding of the effects of ROO.


As a complement to the quantitative method described above, interviews were also conducted in
Madagascar to get a better knowledge of the Malagasy textile and clothing sector. Interviews
were held with representatives for private sector organizations, companies and the Delegation of
the European Commission in Madagascar. It is important to note that this is only a case study
and that the interviewed companies and organizations do not represent a scientific sample of the
textile and clothing sector in Madagascar. The current political situation in Madagascar also
forced me to shorten my stay in Antananarivo and made the number of visited companies quite
small.


All material, except interviews, in this paper is collected from articles published in scientific
journals, reports from leading organizations/research centers like the World Bank or from
electronic sources. Electronic sources are foremost taken from the sites of governments and such
organizations just mentioned to get a high reliability. Data has been collected from different
well-known sources, see above, and then compared to make sure that a just result is presented.



1.4 Disposition
Chapter two of this paper gives a quick background to the textile and clothing sector in
Madagascar. Special attention is devoted to the EPZ’s importance for the sector’s development

                                                                                                 9
and the effects of the phasing out of the Multi Fiber Agreement (MFA). The next chapter covers
the economic theory in relation to ROO to give a theoretical background to the examined issues.
Chapter four further describes the different preferential trading arrangements of the EU and the
US, Lomé/Cotonou and the African Growth and Opportunity Act (AGOA). This is followed by
chapter five which presents the empirical material of the paper. This part is focused on the
Malagasy export and import responses to Lomé/Cotonou and AGOA and these responses’
connection to ROO. Finally, chapter six summarizes the study’s findings and gives an update to
the latest developments in the Malagasy textile and clothing sector.




1.5 Previous Research
A few studies have previously examined the impact of ROO on clothing exports from sub-
Saharan Africa. The entry into force of AGOA triggered several comparative studies on the
effects of the EU and US preferential trading arrangements for LDCs. Most of the studies
mentioned below looked at sub-Saharan Africa as a group and Madagascar was included in a
majority of the samples. None of the studies is a case study of only Madagascar.


That strict ROO have a negative effect on preferential market access is generally accepted and
supported by findings by Cadot, Djiofack and De Melo.6 When it comes to African clothing
products, it is foremost the limitations in fabric sourcing that reduces the effects of tariff
concessions. Portugal-Perez has investigated this further and found that the relaxing of ROO by
allowing producers to use fabric from anywhere in the world would increase African clothing
exports by as much as 300 %. Rules of origin have also been found to restrict the possibilities of
export diversification.7 De Melo and Portugal-Perez have showed that this restriction of
diversification is expected to be in the 30-60% range.8 In addition, Ahmad has concluded that
liberal ROO lead to increased sourcing from developing countries which means that strict ROO
serve as impediments to South-South trade. The cost of this limited South-South trade as an
effect of strict ROO is estimated to be USD 20 million annually.9 Rules of origin have also been
showed to restrict developing countries when it comes to efficiency and producer networks.


6
  Cadot, Djiofack and De Melo (2008) p. 45
7
  Portugal-Perez (2008) p. 21
8
  De Melo and Portugal-Perez (2008) p. 19
9
  Ahmad (2007) p. 34
                                                                                               10
Brenton and Ozden found that more liberal ROO would increase the African textile and clothing
producers’ possibilities to create an efficient industry through enhanced integration into global
and regional production networks.10 Finally, PricewaterhouseCoopers conducted a case study of
the garment sector in Lesotho on behalf of the European Commission in relation to the Economic
Partnership Agreement (EPA) negotiations. The strict preferential ROO were found to be the
core reason for the low volume of Southern African Development Community (SADC) clothing
exports to the EU. Increased fabric sourcing flexibility through a single transformation rule, as
AGOA ROO, or lower value-added thresholds were recommended.11




10
     Brenton and Ozden (2005) pp. 20, 22
11
     PricewaterhouseCoopers (2006) pp. 37, 72
                                                                                              11
2 Madagascar and Textiles


Chapter two serves as a background chapter and aims to give a better understanding of the
present situation in the Malagasy textile and clothing industry. The first part is a sketch of the
economic background of the country, which is followed by an introduction of the Malagasy T&C
sector. A special focus is diverted to the EPZ’s and the Multi Fiber Agreement’s importance for
Malagasy clothing exports.



2.1 Madagascar’s Economic Background
In the first years of independence from the French colonial rule in the 1960s, Madagascar was
one of the better-off countries in sub-Saharan Africa when it came to income and living
standards. This head-start was later lost due to decades of economic mismanagement during
which per capita income declined, from USD 473 in 1970 to USD 290 in 2005.12 In the 1980s,
the government started to implement reforms supported by the International Monetary Fund
(IMF) and the World Bank, in an attempt to get the country’s economic situation back on track.
Import substitution was abandoned in favor of a more outward-looking development strategy and
the allocation of public resources was improved.13 In the mid 1990s, the economic effects of the
reforms finally started to materialize in terms of higher Gross Domestic Product (GDP) growth
and increased exports. So far the economic improvements have however only reached a few
geographical areas due to inadequate redistribution policies.14 Madagascar is today one of the
poorest countries in the world, and more than two thirds of the population live below the poverty
line of USD 1 a day. The country has a Human Development Index (HDI) of 0.533 which ranks
the country as number 143rd out of 177 countries. The ranking places Madagascar as a country of
medium human development, surrounded by countries like Nepal and Cameroon.15




12
   http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/AFRICAEXT/MADAGASCAREXTN/0,,menuP
K:356362~pagePK:141132~piPK:141107~theSitePK:356352,00.html
13
    Cling, Razafindrakoto and Roubaud (2004) p. 5
14
    AfDB/OECD (2005) p. 279
15
    http://hdr.undp.org/en/statistics/
                                                                                               12
Figure 2.1 shows the annual GDP growth from 1990 to 2006. The overall positive picture is only
disrupted by the negative growth rates of 1991 and 2002. That economic progress was put to a
halt these years is explained by severe political instability. The big drop in the growth rate from
2003 to 2004 is partly explained by the two violent cyclones that hit Madagascar in the
beginning of 2004.16 Cyclones are a recurrent problem every rain season and often have severe
consequences for the already poor Malagasy population.

Figure 2.1 - Madagascar’s annual GDP growth in % 1990-2006
     %
     15


     10


      5


      0
           1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
      -5


     -10


     -15


Source: WDI Online




2.2 The Textile and Clothing Sector
The labor-intensive textile and clothing sector is a key sector for many developing countries. It is
a sector, possibly the only successful one, where poor countries have been able to diversify and
increase exports. This has been possible through exploitation of developing countries’
comparative advantage in low-cost labor. Further characteristics that make the T&C sector
suitable for developing countries are low start-up investments, simple technology, a demand for
low-skilled labor and limited importance of scale economies. That many LDCs in sub-Saharan
Africa have preferential access to the EU and US markets has provided an additional incentive
for developing T&C production in this area of the world.17


Madagascar’s T&C industry is one of the fastest expanding in sub-Saharan Africa.18 The sector
has gone through a period of impressive growth since the government’s more liberal economic

16
   AfDB/OECD (2005) p. 280
17
   Brenton and Ozden (2005) p. 4
18
   Nicita (2006) p. 5
                                                                                                  13
policies were introduced in the late 1980s. The sector’s expansion has been especially positive
during the last 15 years after the creation of the EPZ. In 1990, the number of garment producing
firms was only 10.19 In 2007, this number had increased to 120. By then the Malagasy garment
manufacturers employed 120 000 people, making the T&C sector supply over 30% of formal
jobs in the economy.20 The textile and clothing sector’s focus on export has made clothing the
dominant export good today, as can be seen in Table 2.1 below.

Table 2.1 - Top 5 exported products in 200621

                                                                                                                    USD million           % of total export
1. Articles of apparel and clothing accessories                                                                                    583                 49,9
2. Fish, crustacean and molluscs, and preparations thereof                                                                         218                 18,6
3. Coffee, tea, cocoa, spices, and manufactures thereof                                                                            133                 11,4
4. Vegetables and fruit                                                                                                             40                  3,4
5. Miscellaneous manufactured articles, nes                                                                                         23                  1,9
Other commodities                                                                                                                  172                 14,7
Total                                                                                                                             1169                  100
Source: UN COMTRADE



Figure 2.2 – Textile and clothing exports as a share of total export22
      %
      60

      50

      40

                                                                                                                                    Clothing
      30
                                                                                                                                    Textiles
      20

      10

       0
           1990

                  1991

                         1992

                                1993

                                       1994

                                              1995

                                                     1996

                                                            1997

                                                                   1998

                                                                          1999

                                                                                 2000

                                                                                        2001

                                                                                               2002

                                                                                                      2003

                                                                                                             2004

                                                                                                                    2005

                                                                                                                           2006




Source: UN COMTRADE



Further, Figure 2.2 shows the evolution of textile and clothing exports as shares of total export.
From having had a negligent share of exports in the beginning of the 1990s, clothing alone now

19
   http://www.mefb.gov.mg/don_stat/don_sec/t12_nbzfis.pdf
20
   Global Development Solutions (2007) p. 21
21
   The products are classified according to SITC Rev. 2 where, Articles of apparel and clothing accessories have
number 84, Fish, crustacean and molluscs, and preparation thereof number 3, Coffee, tea, cocoa, spices &
manufacs. thereof number 7, Vegetables and fruit number 5 and Miscellaneous manufactured articles, nes number
89.
22
   Exports are mirror exports.
                                                                                                                                                              14
accounts for 50% of total exports. Textile has on the other hand had a rather stagnant
development and has therefore not contributed to the sector’s expansion. In addition, the Figure
shows that exports of clothing were hit hard by the political crisis in 2001-2002 but also that the
sector recovered quickly and now exports more than before the crisis year of 2002. Finally, it
seems like the fast expansion pace of clothing exports has experienced a slowdown since 2005
which could be explained by the phasing out of the Multi Fiber Agreement, see section 2.2.2 for
more information.


The main destinations for Malagasy textile and clothing exports are the EU and the US. Europe,
and in particular France, has traditionally been the most important trading partner for
Madagascar. The entry into force of AGOA in 2000 has however changed the picture and the US
is now also a major export destination. In the year 2006, the EU and the US received 52% and
43%, respectively, of Madagascar’s total T&C exports. The export to the US has showed a
slowdown the last few years, while the export to the EU on the other hand has been increasing
slightly.23 Most of the clothing exports go to specialized retailer chains, supermarkets and mail
order firms like the GAP, Carrefour and La Redoute.24



2.2.1 Export Processing Zone
The Export Processing Zone, or Zone Franche, was introduced in 1990 as a part of the export-led
growth strategy under the structural adjustment program of the IMF and the World Bank. Law
91.020, passed a year later in 1991, defines the rules for the free-zone companies.25 The main
eligibility requirement for EPZ participation is export orientation. Companies must export at
least 95% of their production or provide services and/or inputs to EPZ exporters. In addition,
employment opportunities must be created and adequate environmental safeguards must be
provided. Companies should also “strive to achieve significant technical know-how and
technological transfer”.26


The rewards for meeting the requirements are found in several tax breaks and special regulations
concerning foreign ownership. EPZ companies are exempt from all duties and taxes on both

23
   COMTRADE and author’s calculations
24
   Limantour (2006) p. 12
25
   Cling, Razafindrakoto and Roubaud (2004) p. 5
26
   Cadot and Nasir (2001) p. 6
                                                                                                15
exports and imports. However, in order to hinder firms supplying the local market from setting
up an EPZ company, a Value Added Tax (VAT) on imported inputs was introduced in 1997.
This VAT can later be refunded if the company shows a proof of export of the final good for
which the input was imported. There is also a total exemption from profit tax the first 2 years for
labor-intensive farming and fishing companies, and the first 4 years for industrial and service
companies (i.e. T&C companies). After the first grace period, companies pay a fixed 10% in
profit tax which is substantially lower than the normal 35% for non-EPZ companies. Moreover,
companies are given a profit tax reduction equal to 75% of the cost of new investment. Important
is also the special access to foreign currency and the total freedom for capital transfers for all
EPZ companies.27 A 100% foreign ownership is further allowed as well as free repatriation of
profits after payment of taxes. Since foreigners could not get property rights to land at the time
of the passing of the EPZ law, foreign companies were granted 99 year leases for investment in
land.28


Thanks to the EPZ, Madagascar has experienced an export boom from the mid-1990s and has
become the only successful African new exporter of manufactures, in addition to Lesotho, during
the last decade. The free-zone has hence helped Madagascar to move from a dependency on
agricultural products, mainly vanilla and coffee, to a more diversified economy. The
manufactures produced mainly belong to the T&C sector. In 2001 Madagascar became the
number two African clothing exporter in Sub-Saharan Africa after Mauritius.29 The same year
the sector accounted for around 90% of total EPZ output in Antananarivo and Antsirabe, where
the majority of EPZ companies are located.30 The T&C sector is still in total dominance of the
EPZ today and accounts for more than 50% of all companies in the Zone.31


The expansion of the EPZ was from the start driven by French investors who were attracted by
the many French-speakers among the Malagasy population and by an already large French
community.32 Investors became more diverse in time among which Mauritian companies have
been especially important for the T&C boom. Mauritius is a country that a few years previous to

27
   Cling, Razafindrakoto and Roubaud (2005) p.787
28
   Cadot and Nasir (2001) p. 6
29
   Cling, Razafindrakoto and Roubaud (2007) p.5
30
   Cadot and Nasir (2001) p. 6
31
   http://www.gefp.com/statistiques.php?id=secteurs
32
   Cling, Razafindrakoto and Roubaud (2005) p.787
                                                                                                16
Madagascar experienced a similar development of their own textile industry. When wages rose
in Mauritius and a US quota restriction on Mauritian imports was introduced, Mauritian
investors looked for a suitable place to relocate. Important reasons why Mauritian investors
chose Madagascar were the proximity to Mauritius, the high productivity of Malagasy workers
compared to other Sub-Saharan countries, the French-speaking population, no quota restrictions
and the low labor cost. It can be mentioned that the average monthly salary for a machine
operator in Madagascar is less than one-third of that in Mauritius. However, managers have
stated that Madagascar would not have been interesting without the implementation of the EPZ
initiative despite the just mentioned advantages.33 A third phase of investment inflow came from
Asia in the late 1990s in relation to the entry into force of AGOA in 2000. Production abroad
was a way to circumvent the textile quotas faced by Asian companies. The choice of Madagascar
was again motivated by the EPZ, low labor costs and by preferential access to both the EU and
the US.34 The most common nationalities among EPZ owners in 2008 were French (29%),
followed by Malagasy (20%), Mauritian (16%) and Chinese (12%).35


In sum, the success of the Malagasy textile and clothing sector has been based on a combination
of factors: low labor costs and a relatively high productivity have given unit production costs
among the lowest in the world, preferential access to the US and EU markets has created a
possibility for foreign investors to circumvent their own restricted access to the large markets, a
French-speaking population has simplified potential communication difficulties and the EPZ has
created the essential incentive to attract the foreign firms.



2.2.2 The Multi Fiber Agreement
Textiles and clothing have traditionally been a heavily protected sector both in the US and in
Europe. The competition from developing countries using their comparative advantage in low-
cost labor was considered to threaten jobs in the labor-intensive T&C sector. Already in the
1950’s, Asian low-cost countries agreed to introduce voluntary export restraints for cotton
textiles to the US. In 1962 the Long Term Agreement Regarding International Trade in Cotton
Textiles (LTA) entered into force, and was later renegotiated and replaced by the MFA in 1974.


33
   Cadot and Nasir (2001) p. 7
34
   Cling, Razafindrakoto and Roubaud (2005) p.787
35
   http://www.gefp.com/statistiques.php?id=nationalites_proprietaires
                                                                                                17
The agreement extended trade restrictions to wool and man-made fibers in addition to cotton.
The MFA aimed at avoiding market disruptions when new markets opened up to trade which
meant that the agreement was supposed to be temporary during an adjustment phase. There was
however no clear definition of what constituted a “market disruption”. Consequently, the MFA
came to comprise most developing country textile exports to the EU and the US. A system of
bilateral quota agreements was set up that violated the principles of the multilateral trading
system in several ways: it used quantitative restrictions instead of tariffs, it violated the most
favored nation principle, it discriminated against developing countries and it was non-
transparent. The MFA finally expired in 1994 but was followed by the Agreement on Textiles
and Clothing (ATC) the purpose of which was to serve as a transitory regime between the MFA
and the complete integration of T&C into the multilateral trading system. The ATC
progressively phased out the quotas during a period of ten years. From 1 January 2005, T&C
have been subject to the general rules in the General Agreement on Tariffs and Trade (GATT),
after 40 years of restricted trade.36


The phasing out of the quotas has brought about changes for Malagasy T&C producers. It has
been shown that there is a strong correlation between the success of export processing zones and
the MFA because the foreign direct investments in the zones have been a way to circumvent the
quotas. It was expected that Asia, and especially China, would be the main beneficiary of the
phasing out of the quotas. This has also turned out to be true. Even if Madagascar as a LDC still
benefits from preferential access to EU and the US, the main incentive for Asian investors,
circumventing the quota, is no longer there.37 This means that less investment from Asia is
expected in the following years. It is also hard for Madagascar to compete with the high
productivity of Asian workers. The increased international competition has already been noticed
in Madagascar. The expansion of the Zone Franche and T&C exports has come to a halt since
2005, see Figure 2.2. This development is far from unique; T&C producers in Africa have
suffered from a decline or stagnation of exports in general after the MFA phase out. Of the major
African T&C exporters (Mauritius, Madagascar, Lesotho and Kenya), Madagascar is still the
country that has had the most positive development since 2003.38 See Table A.1 in Appendix for



36
   Nordås (2004) pp. 1, 13
37
   Cling, Razafindrakoto and Roubaud (2004) p. 22
38
   Cling, Razafindrakoto and Roubaud (2007) p. 8
                                                                                               18
a closer look at the development of African and Asian T&C exports after the phasing out of the
MFA quotas.


Investments did decline after 2005 in Madagascar and some Chinese owned clothing companies
closed down.39 Many investors have however come back the last two years after disappointments
with Asian quality.40 Nonetheless, Madagascar will never be able to compete with the low costs
and weak environmental protection of Asia (foremost China and Bangladesh) when it comes to
basic apparel production in the long run. Madagascar is therefore now trying to specialize in
more technical products with higher value-added like embroidery. Designer brands with haute
couture lines have for example taken an interest in Madagascar lately and this market is deemed
to have a great potential if exploited correctly.41 The new direction of the industry can possibly
explain why Madagascar has had the most positive development of the major African T&C
exports since 2003.


Lastly, the increased competition from Asia has also affected wages and labor standards in the
EPZ. From having been a driving force for better working conditions in Madagascar, the EPZ
has since 2005 lowered wages, labor standards and non-wage benefits. Working hours have
gotten longer, company medical services, paid holidays, etc. which used to be much higher than
in the private formal sector have now been substantially reduced. Wages are nonetheless still
higher on average than in the informal sector, which is the main alternative for the low-skilled
mainly female labor force of the Zone Franche.42




39
   Interview Text’île Mada
40
   Interview GEFP
41
   Interviews Text’île Mada and GEFP
42
   Cling, Razafindrakoto and Roubaud (2007) p. 18
                                                                                               19
3          The Economics of Rules of Origin



This chapter will begin with an introduction to and a definition of ROO. What are the different
types and how is origin determined? Second, the effects of ROO are analysed and the critique
that ROO are a protectionist tool is met. Special attention will be diverted to ROO and
preferential trading arrangements for developing countries.




3.1 Background to Rules of Origin
Two types of ROO exist, non-preferential and preferential. Non-preferential ROO are simply
used to determine where a product is produced. They are a way to separate domestic from
foreign products so that for example safeguard measures and origin marking can be used.
Preferential ROO, on the other hand, are used in a PTA to establish if products exported from
one member to another qualifies for preferential treatment in the form of better market access.43
The better market access usually comes in the form of a lower customs tariff, or in the case of a
Free Trade Agreement (FTA), a tariff-free entry. The rest of the paper will focus on the role of
preferential ROO in FTAs.


The usage of preferential ROO in a FTA is justified by the risk of trade deflection. Trade
deflection is a form of fiscal fraud and occurs when products from non-members are shipped
through a transit country, a FTA-member with a low tariff, to a member country with a higher
tariff.44 This trade deflection is possible since the FTA practises free trade between members but
it has no common external tariff like a customs union has. An example of trade deflection would
be if country A and country B are members of the same FTA and practise free trade between
each other. Country B can then ship its products tariff free to country A and vice versa. Country
C is on the other hand not a member of the FTA. Therefore C has to pay tariffs when exporting
to both A and B. For argument’s sake, we can say that country A has a 10% tariff on all imports


43
     Estevadeordal and Souminen (2003) p. 1
44
     Krishna (2005) p. 3
                                                                                               20
from non-FTA members and country B only a 2% tariff. In that case, country C could
theoretically ship its goods to country B, pay the 2% tariff and then re-export to country A tariff-
free. In this way, the higher tariff in country A is avoided.


Due to the fact that the production process often is fragmentized and conducted in several
different countries it has gotten harder to determine a product’s origin. As a consequence, the
preferential ROO have become rather complicated. There are today different methods of
determining origin of a product. There are two central criteria, recognized by the Kyoto
Convention, to determine origin of a product: wholly obtained or produced and substantial
transformation.45 The wholly obtained or produced criterion is uncontroversial and means that
the country where the product has been entirely grown/harvested/extracted from the soil or
produced using only material that has been domestically grown/harvested/extracted from the soil
is the origin country. Simply put, no second-country components or material is allowed in the
production process.46


The substantial transformation criterion is more difficult and subjective. There are three main
ways to determine if the product has gone through substantial transformation and different tests,
sometimes a combination of tests is necessary to determine origin. The first possible test is a
change in tariff classification according to the Harmonized Tariff Schedule between the input
and the exported good. This can be more or less strict depending on how extensive the change
must be. The change in tariff classification rule can demand that the product alter its chapter (2-
digit level), heading (4-digit level), sub-heading (6-digit level) or item (8-10-digit level). The
change on chapter level is the strictest of these versions. The second test is value content which
means that a certain percentage of value must have been added to the product to get origin status.
This criterion can take three different forms: a) import content test: imported inputs are not
allowed to exceed a certain percentage of the final good’s value; b) domestic content test: a
minimum percentage of local value must be added in the last country where the product was
processed; c) value of parts test: originating parts must account for a certain percentage of the
final good’s value.47 The exact percentage differs between trade agreements and products but for


45
   The Kyoto Convention has been adopted by the World Customs Organization to standardize and harmonize
customs procedures and policies around the world. It was originally adopted in 1974.
46
   Estevadeordal and Souminen (2003) p. 3
47
   Falvey and Reed (1998) p. 213
                                                                                                          21
the import content test the percentage is usually around 40%.48 The third test is a specific process
rule which requires the product to undergo (positive test) or not undergo (negative test) a certain
manufacturing process in the originating country.49


There are advantages and disadvantages with all three tests. The change in tariff classification
test has the advantage of being relatively simple, easy to apply and it could be used uniformly
across countries. Difficulties, on the other hand, are applying a commodity classification
designed for several purposes and the habit of having lists of exceptions when it comes to this
rule. The value content test is potentially quite costly since it requires examination of production
costs, decisions on how to value inputs and what inputs should be used. The advantage of the
specific process rule test is that it can be adjusted to any special case. However, it is very
difficult to construct technical tests for every product. It is also hard to verify information in
order to determine which process actually took place in third countries and therefore it can be
easy to manipulate the facts.50


Finally, there exist some complementary rules, used in combination with one or several of the
above tests, to make ROO less restrictive. Cumulation rules allow producers to use inputs from
certain countries without losing the preferential status given to them in a FTA. This cumulation
can be bilateral, diagonal or full. Bilateral cumulation is the most common form and operates
between two FTA partners. Partner A can use inputs originating from the partner B as if they
were A’s own, and vice versa, without affecting the final good’s origin status. Diagonal
cumulation means that beneficiary countries that are members of the same preference program
can use inputs originating from each other and still be granted origin status for the final product.
Full cumulation is the least restrictive form and extends diagonal cumulation. Here, countries
tied together by the same preferential origin rules can use goods produced in any part of the area,
even if they were not originating products. All processing done in the preference area is regarded
as if it had taken place in the final processing country.51




48
   Augier, Gasiorek and Lai-tong (2004) p. 1452
49
   Estevadeordal and Souminen (2003) p. 4
50
   Falvey and Reed (1998) pp. 213-214
51
   Ahmad (2007) pp. 10-11
                                                                                                 22
Two other complementary rules exist. First, the de minimis rule that accepts a certain maximum
percentage of non-originating inputs in the production without affecting origin. Finally, there is
the absorption principle which means that products which have gained origin status are allowed
to be considered originating when used as inputs in subsequent production.52




3.2 Rules of Origin - a protectionist tool?
Rules of origin can have an impact on trade and investment flows and how well a FTA with
ROO promotes trade liberalization has been questioned. ROO raise domestic production and
administrative costs and this in turn affects trade. Production costs are increased due to supply-
switching effects and the technical criteria imposed by the ROO regime. The supply-switching
leads to trade diversion and the rise in production costs can also decrease final goods production.
Administrative costs are increased because producers have to get the origin certification and
because the import country’s customs has to verify the origin. The administrative costs work as a
form of transaction cost and are likely to reduce bilateral trade creation.53 An attempt to compare
costs between different ROO has been done by Carrère and de Melo. They found that a change
in tariff heading is the least costly to comply with, followed by value content rules and finally, a
technical requirement is found to be the most costly.54 ROO also have an investment effect. In
the long run, ROO can create investment diversion when extra-FTA producers choose to locate
plants inside the FTA in order to satisfy the ROO and benefit from preferential treatment.55



3.2.1 Rules of Origin as transaction costs
The complexity of the origin rules can act as a blocker of the potential trade creating effects of
the tariff concessions granted in a FTA. Trade creation is the phenomenon when domestic
production is replaced by cheaper imports, due to relative price changes when free trade is
introduced, from a more efficient FTA-partner. Different certification mechanisms impose costs
on both firms and governments. These costs are often far from small and increase even more
when countries are members in several FTAs, which is a common phenomenon. The
administrative burden offsets the tariff liberalization and leads to underutilization of preferences.

52
   Estevadeordal and Souminen (2003) pp. 4-5
53
   Augier, Gasiorek and Lai-Tong (2005) p. 576
54
   Carrère and de Melo (2004) p. 27
55
   Estevadeordal and Souminen (2003) p. 8
                                                                                                  23
Firms can simply avoid all the ROO administrative costs if they choose to act as Rest of the
World (ROW) producers and instead pay the non-preferential tariff. As long as this tariff is
reasonably low, which it often is between developed countries, the distortion of trade between
FTA-partners is limited but trade creation will, of course, still be affected.56 In the case of a high
non-preferential tariff, the effect on trade creation is larger and producers can then choose to use
inputs from the ROW and only produce for the home market and the ROW instead of producing
for the FTA partner.




3.2.2 Increased production costs through supply-switching
According to classic Vinerian analysis with trade creation and trade diversion, FTAs will lead to
some supply-switching, with FTA-suppliers as the beneficiaries, when preferential tariffs change
the relative price of imports. Partner country imports become relatively cheaper than before and
third country imports become relatively more expensive, which induces a natural supply-
switching in favour of partner firms despite more efficient third country producers. Supply-
switching to a FTA-supplier from a more efficient third country producer is called trade
diversion and creates economic inefficiencies. This classical analysis does not take ROO into
account which is unfortunate since ROO can aggravate the supply-switching effects of FTAs.


The impact of the preferential tariff aside, ROO distort trade patterns both between members of
the FTA and between members and the rest of the world. In order to meet the ROO, firms often
have to change suppliers from extra-FTA to intra-FTA (to domestic suppliers in the case of no
cumulation). This means that ROO create an extra element of trade diversion. In the ordinary
trade diversion case described above, a country switches suppliers to a partner country because it
would be cheaper when zero tariffs between partners are introduced. In the case of FTAs and
ROO, producers would switch suppliers even if it meant more expensive inputs to comply with
the ROO. This additional trade of inputs between members can easily be mistaken for evidence
of trade creation which they are not. That ROO have a supply-switching effect can be used by
protectionist governments or industry lobbies. A country can protect its own industries on a
partner country’s market even if the partner country in question has zero tariffs to the rest of the



56
     Augier, Gasiorek and Lai-Tong (2005) p. 576
                                                                                                   24
world. A country with high cost producers can in this way export protection of its industries to
member countries through restrictive ROO.57


The effect of supply-switching because of restrictive ROO on production costs can be shown
graphically according to a model developed by Kala Krishna.58 Figure 3.1 shows how costs rise
when the choice of inputs is limited through a value content test with bilateral cumulation. The
model contains two types of inputs, FTA-inputs (L) and third country inputs (K), that are used to
produce the good in question under constant returns to scale. The curve in the figure depicts the
unit isoquant which symbolises the number of goods produced. At a given price for L and K
firms will choose the input mix at point Z, using L and K so that their ratio equals α. The height
of the line AB represents the lowest unit costs attainable for that input mix. If binding ROO then
are introduced which require L/K to be at least α’ > α, Z would no longer be a feasible option
because only points below the ray from the origin with slope 1/α and above the isoquant would
be possible. The unit costs are therefore now minimized at point X and represented by the height
of line DE. The ROO have raised the production costs and distorted the input mix in favour of
the FTA-inputs for any given level of output. It is also possible to see that as the ROO become
more restrictive, as α rises and the ray from the origin swings down, unit cost will rise, shift out.


Figure 3.1 – Value content ROO and Costs

      K




      K

      D                            L/K = α


      A



                                             L/K = α’ > α


                    Z
                               X



          0                    B     E               L




57
     Krueger (1993) pp. 9-10
58
     Krishna (2005) p. 10
                                                                                                   25
3.2.3 Trade preferences for developing countries and ROO
The LDCs have received preference treatment in the multilateral trading system since 1971.
Preferences are granted in the form of reduced or zero tariff rates over the Most Favoured Nation
(MFN) rates for selected products in order to increase LDCs’ export earnings, promote their
industrialization and to accelerate their rates of economic growth. Several programs for trade
preferences for developing countries have evolved in the world, presently there are 13 national
preference schemes notified to the United Nations Conference on Trade and Development
(UNCTAD).59


The motive for including ROO in preferential trading arrangements differs from the usual trade
deflection justification. In the case of North-South integration it would namely be the preference
granter that would benefit from trade deflection since tariffs in general are lower in developed
than in developing countries.       Reasons why ROO are seen as necessary are hence found
elsewhere. First, preference-granters do not want to extend the preferences involuntarily to non-
eligible producers or producers that only transform their goods superficially. Second, the ROO
can be a way to control the process of preferential liberalization through a reduction of
adjustment costs in the North. Third, the usage of ROO in preferential programs is sometimes
described as a tool for development policy because ROO are believed to encourage vertical
integration in developing countries. The demand to use domestic inputs is seen as a form of
infant industry protection of local producers further down the value chain.60


Vertical integration is however not an easy target in developing countries where light and labor
intensive industry often is the only possible and credible industrialization option. It is therefore
harder to create sustainable vertical integration since input industries can be more capital
intensive than the final good production. If vertical integration takes place, the result risks being
inefficient due to the protection effect of the ROO which means that the new input-supplying
industries serve to institutionalize the trade deflection effect of regional integration.61 Certain
sectors’ characteristics make it especially hard to make ROO lead to an increase of locally
sourced inputs or vertical integration. This is for example true for the textile and clothing


59
   http://www.unctad.org/Templates/Page.asp?intItemID=2309&lang=1
60
   Cadot, Djiofack and De Melo (2008) p. 13
61
   Cadot, Djiofack and De Melo (2008) p. 14
                                                                                                  26
industry where global dynamics require unregulated access to low cost materials to be
competitive and where “buyer-driven value chain realities” dictate producer choices.62


Restrictive ROO can lead to strong supply-switching effects that are more damaging in
developing countries just because they often lack domestic suppliers of inputs and because it is
hard to build up such an input industry. These effects can be shown in a sketch of the hub-and-
spoke system which exemplifies how both FTA-members and third countries are affected by
ROO. A hub-and-spoke system evolves when one country forms bilateral FTAs with many other
countries, for example when the EU creates a preferential trading arrangement for developing
countries. The EU then becomes the “hub” of the system and the developing countries that are
connected to the EU through bilateral FTAs are called “spokes”. Below, Figure 3.2 shows a hub-
and-spoke pattern. Bilateral FTAs, as mentioned above, tend to diminish trade between spokes
because of two phenomena: the preferential tariffs the spokes grant firms in the hub but not firms
in other spokes, and the ROO that lead to supply-switching towards domestic or partner firms in
order to meet origin requirements. Exports from the ROW to the spokes also tend to be
depressed because of the same reasons just mentioned. There is however no first-order effect on
exports from spokes to the ROW.63 Diagonal or full cumulation can of course reduce the loss of
exports between spokes.


Figure 3.2 – Hub-and-spoke system


                                              spoke
                                                                            RoW
                                                          spoke             Reduction in exports from RoW to spoke
                                  spoke
                                                                  Reduction in exports between spokes

                             spoke             Hub             spoke
                                                        Increase in exports from Hub to spoke


                                   spoke                   spoke
                                              spoke




62
     Neumann (2008) p. 11
63
     Augier, Gasiorek and Lai-Tong (2005) pp. 579-580
                                                                                                             27
If a developing country takes part in a preferential trading arrangement and there are no domestic
suppliers of inputs, the country will be forced to use expensive inputs from the hub instead of
using the cheapest option on the world market in order to comply with the ROO. In this way it
will be hard for the developing country to develop its industries and to compete on the world
market since the products produced will be more expensive and more inefficient than otherwise
would have been the case. It is also harder for a poor country with low labor costs to comply
with high value-added requirements in the ROO relative to countries with higher labor costs.64
Moreover, it happens that the ROO themselves are so strict for labor intensive products, which is
what developing countries in general export, that it is difficult for spokes to gain origin status for
this type of goods.65 This means that the value of the tariff concessions is greatly reduced for
developing countries and that the preferential trading arrangements can be more protectionist
than first thought.




64
     Brenton (2006) p. 281
65
     Brenton and Manchin (2003) p. 767
                                                                                                   28
4        Preferential Trading Arrangements


Chapter four presents the EU and the US preferential trading arrangements for LDCs with a
focus on the ROO of the programs. Madagascar is eligible for preferential market access to both
the EU and the US market, a fact that has largely contributed to the development of the
Malagasy T&C sector. Since Madagascar is an African country, only preference programs
concerning Africa will be covered below.




4.1 EU Preferential Trading Arrangements
4.1.1 Lomé and Cotonou
Europe’s colonial history has led to a special cooperation between the EU and the African,
Caribbean and Pacific (ACP) countries. The beginning of this cooperation dates back to the
Treaty of Rome establishing the European Economic Community (EEC) in 1957, then with a
particular focus on French-speaking African countries. The adhesion of the United Kingdom to
the EEC in 1973 started a wider reaching cooperation program. In 1975 the first Lomé
convention governing trade relations between the EEC and 46 ACP countries was signed in
order to promote economic development. Several updates of the convention followed but the
main focus of non-reciprocal trade preferences, mainly duty-free and quota-free access, for most
ACP exports to the EU remained the same.66


The Lomé convention was replaced by the Cotonou Agreement in 2000 to prepare the EU-ACP
collaboration for complete fulfilment of the World Trade Organization (WTO) rules since non-
reciprocal exclusive trade preferences violate the principle of non-discrimination in the GATT.67
The old trade preference scheme remained in force until 1 January 2008 when EPAs replaced the
trade chapters of the Cotonou Agreement. The EPAs entail a deeper cooperation with rights and
obligations for both sides of the partnership, a partnership based on reciprocity.68 Each EPA is

66
   http://ec.europa.eu/development/geographical/cotonou/lomegen/lomeitoiv_en.cfm
67
   http://www.acp-eu-trade.org/index.php?loc=epa/background.php
68
   http://ec.europa.eu/trade/issues/bilateral/regions/acp/nepa_en.htm
                                                                                              29
negotiated separately between the EU and the country or groupings of countries in question
which means that the degree of trade liberalization will differ between EPAs. ACP-countries that
chose not to sign an EPA before the end of 2007 were offered the Everything But Arms (EBA)
trade initiative instead, an arrangement known to be stricter in terms of market access than ACP-
regulation.69 An important note is that the EBA only was an option for the LDCs among the
ACP-countries. There is no time limit for the EPAs while the Cotonou Agreement was
concluded for a twenty-year period and is hence valid until 2020.70


Madagascar signed an Interim Economic Partnership Agreement (IEPA) as a part of the Eastern
and Southern Africa (ESA) grouping in December 2007. The IEPA entered into force on January
1 2008 and is limited in duration until the full EPA comes into force. This interim agreement was
put in place in order to give more time for finalizing the EPA negotiations which broke down
due to differences in market access offers.71 The IEPA gives Madagascar 100% free access to the
EU market with a transition period for sugar and rice. In exchange Madagascar will
progressively liberalize market access for about 80% of the country’s imports from the EU
during an adjustment period of 15 years. This means that by 2022 80% of Malagasy imports
from the EU will enter the country duty free. Madagascar has also chosen to exclude a range of
sensitive agricultural and industry products from liberalization, for example fish, cereals and
plastic.72 The EPA negotiations are hoped to be concluded before the end of this year but
according to a representative for the European Commission’s Delegation to Madagascar this goal
will probably not be met due to disagreements over levels of development assistance.73




4.1.2 Rules of Origin for Textiles and Clothing
The preferential market access granted to ACP-countries through the Lomé convention and the
Cotonou Agreement was contingent on the fulfilment of certain product specific ROO. The EU
ROO for the T&C sector are known to be quite strict. Clothing has to be manufactured from

69
    http://trade.ec.europa.eu/doclib/docs/2008/march/tradoc_138457.pdf
Everything But Arms is the EU’s Generalized System of Preferences (GSP) from 2001 which provides quota-free
and duty-free access to the EU market for the 50 least developed countries in the world. The agreement covers all
goods but arms and ammunition.
70
   http://ec.europa.eu/development/geographical/cotonouintro_en.cfm?CFID=2311138&CFTOKEN=de5549ec566e5
3bc-44BE1EAC-BCAD-6AE3-85FE869240E498A7&jsessionid=243062fb88384a375d62
71
    Watson (2007) p. 6
72
    http://trade.ec.europa.eu/doclib/docs/2007/november/tradoc_136959.pdf
73
    Interview Delegation of the European Commission in Madagascar
                                                                                                              30
wholly produced or qualifying yarn. The production from yarn further has to comply with the
double transformation rule which means that two different production processes must take place
in the beneficiary country. For clothing this implies that yarn must be woven into fabric which
then, in turn, must be cut and made-up into clothing. Bilateral and full cumulation from African
countries is allowed which gives that qualifying yarn must originate in the beneficiary country,
the EU or any other African ACP-country. It is the full cumulation rules of the Cotonou
Agreement that made most African countries also eligible for EBA continue to export under
Cotonou when the EBA initiative was introduced in 2001. EBA only allows regional cumulation
for the Association of South East Asian Nations (ASEAN), the Central American Common
Market (CACM), and the Andean Community. Hence, African countries are only allowed to use
bilateral cumulation under EBA, meaning that inputs must originate in the beneficiary country or
the EU in order not to lose origin status of the final product.74 Madagascar is one of the many
countries that chose to continue to export under Cotonou even if the country is eligible for EBA.


The IEPA entails substantial changes of the ROO for T&C compared to previous rules in Lomé
and Cotonou. The EU has left the double transformation rule behind and has agreed that only
single transformation is necessary to obtain origin status. This means that Madagascar as a
signatory to an IEPA now is allowed to source fabrics from anywhere in the world, transform
them and then export both quota and tariff free to the EU.75 The ROO for T&C of the final EPA
will not differ from the ROO of the interim agreement.76




4.2 US Preferential Trading Arrangement
4.2.1 African Growth and Opportunity Act (AGOA)
The African Growth and Opportunity Act is part of the Trade and Development Act, a law
signed by President Clinton in 2000. The AGOA initiative provides quota- and duty-free access
to the US market for 38 Sub-Saharan countries.77 The Act originally covered eight years, i.e. the
period from 2000 to 2008, but was extended in 2004 until 2015. Products included for
preferential market access are all goods previously eligible for the US GSP program as well as

74
   De Melo and Portugal-Perez (2008) p. 4
75
   Naumann (2008) p. 8
76
   Interview Delegation of the European Commission in Madagascar
77
   Brenton and Ozden (2005) p. 6
                                                                                               31
an additional 1800 product tariff lines specially added for AGOA. Among the new products are
for example apparel, footwear, wine, chemicals and steel.78


Country eligibility for AGOA takes into account several requirements, including political and
economic criteria. Market-based economies, rule of law, elimination of barriers to US trade and
investment and a strategy for poverty reduction are examples of criteria that must be fulfilled.
Countries that are not eligible at present are Zimbabwe and Sudan. Moreover, the Central
African Republic and Eritrea were removed as AGOA beneficiaries in 2003 while Mauritania
and Côte d’Ivoire have been suspended. Madagascar has been an AGOA beneficiary since the
beginning of the program in October 2000.79


However, AGOA eligibility does not imply the right to export clothing and certain textile
products duty-free to the US. In order to do so countries must, in addition to AGOA eligibility,
also comply with the “Wearing Apparel” provisions. To be qualified for these extra provisions,
countries must implement a product visa system in compliance with the AGOA ROO that has
been approved by US authorities. Presently, 29 of the 38 AGOA eligible countries have fulfilled
the “Wearing Apparel” conditions. Madagascar has been one of these countries since March
2001 and is thereby granted the right to export apparel to the US duty-free.80




4.2.2 Rules of Origin for Textiles and Clothing
If the “Wearing apparel” provisions are fulfilled, countries that want to export apparel duty-free
to the US must also comply with the special set of ROO for T&C. Clothing made in qualifying
countries from US fabric, yarn and/or thread are granted duty- and quota-free access to the US
market with no further limitations. Clothing made from domestically produced fabric and yarn,
or from fabric and yarn produced in other AGOA-beneficiaries has duty-free access to the US
but is subject to a cap of 1.5% of total US apparel imports. The percentage was to increase
progressively to 3.5% of total imports in 2008 and then remain unchanged at 3.5% until 2015.
Amendments to AGOA have however been made that doubled the applicable percentages of this



78
   http://www.agoa.info/index.php?view=about&story=about
79
   http://www.agoa.info/index.php?view=about&story=country_eligibility
80
   http://www.agoa.info/index.php?view=about&story=country_eligibility
                                                                                               32
cap which means that the quota presently is set to 7%.81 The quota has so far not hindered export
from AGOA-beneficiaries. The quota fill rate reached about 60% in 2001-2002 but has after that
been around 30-35% each year.82


There are also additional rules concerning apparel that is mainly made from qualifying inputs. A
garment otherwise eligible for AGOA preferential access shall not be disqualified due to the fact
that the article “contains certain interlinings of foreign origin”. The value of these interlinings
must nevertheless not exceed 25% of the cost of all components of the garment. Finally a de
minimis rule says that apparel can always benefit from duty-free access if the total weight of
foreign, i.e. not of US or AGOA-beneficiary origin, fibers or yarn is not more than 10% of the
total weight of the article.83


What separates AGOA from all other preferential trading agreements is the possibility for LDCs
to benefit from special ROO. LDCs are allowed to used third country fabrics, not US or AGOA,
and still be granted duty-free access to the US market. This means that inputs can be sourced
from wherever in the world.84 The special provisions for LDCs were originally meant to expire
in 2004 but have been extended two times, first until 2007 and now until 2012. All AGOA-
beneficiaries, except Gabon, Mauritius, the Seychelles and South Africa, have enjoyed LDC
status from the start. Mauritius has however benefited from access to third country fabrics since
October 2008 after legislative amendments to AGOA.85




81
   http://www.agoa.info/index.php?view=about&story=apparel_rules
82
   http://www.agoa.info/index.php?view=trade_stats&story=apparel_quotas
83
   http://www.agoa.info/index.php?view=about&story=apparel_rules
84
   Brenton and Ozden (2005) p.7
85
   http://www.agoa.info/index.php?view=trade_stats&story=apparel_trade
                                                                                                33
5      Impact of AGOA and Lomé/Cotonou


This chapter investigates the trade impacts of the US and EU preferential trading arrangements
on the Malagasy textile and clothing industry. The analysis will compare Madagascar’s trade
with the US to the country’s trade with the EU in order to assess the impact of ROO on trade
patterns. In the first step of the analysis, export data is examined and related to AGOA and
Lomé/Cotonou. The second step is devoted to import data since theory predicts that ROO can
affect the choice of input goods. This will also give indications of the importance of the 3rd
country fabric-rule for increasing investments and exports. The chapter ends with an evaluation
of vertical integration and of the utilization of trade preferences.




5.1 Response to AGOA
The figure below shows clothing exports to the EU and the US from the creation of the export
processing zone in 1990 and onwards. This picture gives the starting point of the analysis and
shows the dramatic expansion of Malagasy clothing exports to its main markets, the EU and the
US.


Figure 5.1 - Clothing exports to the EU and the US
      USD million
       400
       350
       300
       250
                                                                                                                                     EU
       200
                                                                                                                                     US
       150
       100
        50
          0
              1990

                     1991

                            1992

                                   1993

                                          1994

                                                 1995

                                                        1996

                                                               1997

                                                                      1998

                                                                             1999

                                                                                    2000

                                                                                           2001

                                                                                                  2002

                                                                                                         2003

                                                                                                                2004

                                                                                                                       2005

                                                                                                                              2006




Source: COMTRADE




                                                                                                                                          34
As Figure 5.1 shows, clothing exports to the US were basically non-existent until the late 1990s.
Interestingly, exports started to increase already in 1998-1999, that is before the entry into force
of AGOA in 2000. Clothing exports increased for example by as much as 136% between 1999
and 2000. This anticipation effect can be a sign of the high expectations investors had of the US
agreement. In 2000-2001, after the entry into force of AGOA, clothing exports to the US
continued to increase rapidly until the crisis year of 2002 when exports plummeted due to
political instability. However, exports recovered very quickly and reached a new all time high
already in 2003. The following year, 2004, even surpassed the numbers of 2003 and is still
today the peak year of Malagasy clothing exports to the US. Exports have namely slowed down,
as Figure 5.1, shows after 2004. This can be related to the already mentioned end of export
quotas for Asian countries in 2005. That exports to the US seem to have been affected more by
the phasing out of export quotas than exports to the EU can in part be explained by the fact that
the export boom to the US has been driven by Asian investors.86 The latest data from OTEXA
show a slight increase in clothing exports to the US again in 2007 to USD 289 million, indicating
that the negative trend might be stoppable.87


These new numbers also support the already mentioned statement by a representative for a
private sector organization in Antananarivo. It seems to be true that Asian companies, who
mainly export to the US, left at first in 2005 after the MFA phase out but then many came back a
few years later.88 Another factor that probably reduced the negative export trend to the US is the
American imposition of safeguard quotas on textile and apparel imports from China in
November 2005. These safeguard quotas did not expire until December 31 2008 so the real
effects of the MFA phase out will not be felt until this year. What this will mean for the
Malagasy T&C sector is yet too early to say but the US has already noticed a growth in imports
from China of certain key clothing products of great importance for Africa.89 It is therefore safe
to say that competition from Chinese producers is likely to get fiercer for the Malagasy T&C
sector in the coming years.




86
   Cling, Razafindrakoto and Roubaud (2005) p. 787
87
   OTEXA
88
   Interview GEFP
89
   Rydberg (2009) p. 2
                                                                                                 35
Figure 5.2 - Exports of Clothing to the US
USD million
  400

  350

  300

  250

  200

  150

  100

   50

    0
        1997   1998      1999       2000       2001       2002       2003       2004        2005       2006


               848 Articles of apparel, clothing accessories, non-textile, headgear

               847 Clothing accessories, of textile fabrics, nes
               846 Undergarments, knitted or crocheted

               845 Outerwear knitted or crocheted, not elastic nor rubberized

               844 Undergarments of textile fabrics, not knitted or crocheted

                    om
               843 W en's, girls' and infants' outerwear of textile fibres, not knitted or crocheted

               842 Men's and boys' outerwear of textile fabrics, not knitted or crocheted

  Source: COMTRADE


To further assess the impact of AGOA, clothing exports to the US have been broken down to a 3
digit level in Figure 5.2. It is possible to see that the product group that has benefited the most
from AGOA is Outwear, knitted or crocheted (845). Yet, the importance of this product group
has decreased after the MFA phase out. In 2006 it was instead Women’s, girls’ and infants’
outwear of textile fabrics, not knitted or crocheted (843) that was the main product group
exported to the US. In general, knitted or crocheted products were more important before the
MFA phase out than after. The last few years, products of textile fabrics that are not knitted or
crocheted have surpassed knitted and crocheted products in export value.


In order to get a general idea of what type of inputs are used in the different clothing products, a
further look at the most exported clothing categories in 1998, 2003 and 2006 is necessary. The
year of 1998 is chosen because it shows the situation before the entry into force of AGOA. The
year 2003 then gives an indication of the trade pattern in the middle of the expansion phase. It is
important here to keep in mind that 2003 was the year after the crisis in 2002. The numbers for
2003 are therefore smaller than otherwise would have been the case. Finally 2006 is the last year
of available data and a year after the MFA phase out.


                                                                                                              36
Table 5.1 - Most exported clothing categories to the US in USD million

                                                                               1998       2003      2006
     8439   Women's, girls' and infants' other outer garments of                 3.6       62.3      85.6
            textile fabrics, not knitted or crocheted
            of cotton                                                          100%        97%        95%
            of man-made fibers                                                   0%         1%         4%
     8423   Men's and boys' trousers, breeches and the like,                     5.6      21.7       52.7
            not knitted or crocheted
            of cotton                                                          100%        98%        74%
            of other fibers                                                      0%         0%        17%
            of man-made fibers                                                   0%         2%         8%
     8451   Jerseys, pullovers, slipovers, cardigans, etc. knitted               5.4      72.2       49.1
            or crocheted
            of cotton                                                           53%        59%        67%
            of synthetic fibers                                                  0%        13%        18%
            of wool or fine animal hair                                         47%         n.a.       n.a.
     8459   Other outer garments and clothing accessories,                       0.4      31.3       27.6
            knitted or crocheted
            of cotton                                                          100%        62%        72%
            of synthetic fibers                                                  0%        37%        28%
     8462   Undergarments, knitted or crocheted                                  1.6      12.7       26.4
            of cotton                                                          100%       100%       100%
Source: COMTRADE



Table 5.1 is organized according to the most exported products in 2006. All three years have the
same most exported products but the order between the products changes during the examined
period.90 Men's and boys' trousers, breeches and the like, not knitted or crocheted (8423) was the
most exported product in 1998, while Jerseys, pullovers, slipovers, cardigans, etc. knitted or
crocheted (8451) was the most exported good to the US in 2003. This category was however not
far behind the most exported one in 1998. In 2006, it is instead Women's, girls' and infants' other
outer garments of textile fabrics, not knitted or crocheted (8439) that has taken over as the most
exported good. The distance to the second most exported good is also larger than before.


Certain trends can also be identified when it comes to the inputs used. Cotton is by far the most
used input during the examined period even if cotton is less important in 2006 than in 1998.
There is also a tendency to use more synthetic fibers and more other fibers in the 2000s in
comparison to the late 1990s.91 Wool or fine animal hair is on the other hand less used in the end

90
   The only exception to this statement is category 8459 which was not one of the top five exported clothing
products to the US in 1998.
91
   Clothes made of synthetic fibers are also subject to a higher tariff than clothes made of cotton when entering the
US. Since Madagascar can export tariff-free to the US it is more profitable to export clothes made of synthetic fibers
than clothes made of cotton. Source: Interview GEFP
                                                                                                                   37
of the period and is no longer one of the most important inputs for products aimed at the US
market.




5.2 Exports to the EU
Because the Lomé convention was in place before the creation of the export processing zone,
that is to say before any large scale clothing production took place in Madagascar, it is harder to
see a direct effect of the EU preferential trading agreement. It is still possible to analyze the
trade pattern to the EU in order to compare possible effects of different ROO.


As both Figures 5.1 and 5.3 show, exports to the EU took off before the export boom to the US.
This is not surprising given the preferential access to the EU market and the historically close
ties to Europe, in particular France. As already mentioned, French investors were also the first to
devote attention to the Malagasy export processing zone. Even if the clothing exports to Europe
expanded relatively fast, the expansion pace must be deemed as slower than the one of exports to
the US in the late 1990s. Further, signs of stagnation for exports to the EU are found already
before the crisis year of 2002. Export values are basically the same from 1999 to 2001. The
exports to the EU also recovered more slowly than those to the US after the crisis. The same
export level as before the crisis year was not reached until 2006. On the other hand, exports to
the EU do not seem to be affected to the same extent as the exports to the US by the end of
export quotas for Asian countries. Instead, export volumes are still growing and the EU is
therefore again the most important export market for Malagasy clothing goods with an export
value of USD 304 million in 2006, see Figure 5.1.


Among the managers interviewed there were diverging opinions about exporting to the EU. One
manager, of a company that mainly exported to the EU, stated that exporting to the EU market
felt more reliable just because European clients tend to react slower to demand changes than the
American ones. To export to the US is therefore sometimes seen as riskier since those exports
are more volatile. In this manager’s view, the slower expansion pace for EU exports was
therefore not that problematic but instead a sign of trustworthy long-term business
relationships.92 On the other hand, a manager whose company mainly exported to the US thought

92
     Interview ARAWAK
                                                                                                38
it was difficult to export to the EU. It was believed to be easier to enter the US market than the
European one, especially because Europeans tended to take slower decisions. At this company,
the faster pace of American decision-making was appreciated and the American business culture
was seen as more transparent.93


Figure 5.3 - Exports of Clothing to the EU
     USD million
       350,00
       300,00
       250,00
       200,00
       150,00
       100,00
        50,00
         0,00
                1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

                     848 Articles of apparel, clothing accessories, non-textile, headgear
                     847 Clothing accessories, of textile fabrics, nes
                     846 Undergarments, knitted or crocheted
                     845 Outerwear knitted or crocheted, not elastic nor rubberized
                     844 Undergarments of textile fabrics, not knitted or crocheted
                     843 Women's, girls' and infants' outerwear of textile fibres, not knitted or crocheted
                     842 Men's and boys' outerwear of textile fabrics, not knitted or crocheted

Source: COMTRADE



Figure 5.3 shows the development of clothing exports to the EU on a 3 digit level to further
compare exports to the EU to those destined for the US. The most exported clothing product
group to the EU from 1993 and onwards is Outerwear knitted or crocheted (845). The
dominance of this product group is even greater than in the US case since it does not show a
slowdown after 2004. It has in fact gotten even more important the last few years. This implies
that exports to the EU do not show any decreased percentages of knitted or crocheted goods, as
was the case for exports to the US. The expansion of exports of Outerwear knitted or crocheted
(845) further means that non-knitted or crocheted products have become less important for the
EU market. The difference between knitted and non-knitted products is interesting and consistent
with the findings of Brenton and Ozden who found that ROO requirements are more costly for
non-knitted than knitted products. This because the ROO for non-knitted items demand that both

93
     Interview MIW
                                                                                                              39
the fabric and the yarn of which the fabric is made must come from the EU, ACP-countries or
Madagascar. Since there usually is no fabric involved in the making of knitted products the ROO
become less costly to meet in that case.94 That Madagascar exports more knitted products to the
EU than to the US can hence be an effect of the different ROO because it is more costly to
produce non-knitted products for the EU than for the US.


A look at the top five exported clothing goods to the EU in 1994, 2000 and 2006 in Table 5.2
also shows a different composition of the most important clothing categories in comparison to
the US case. Clothing accessories of textile fabrics, not knitted or crocheted (8471) was the most
exported good to the EU in 1994 while Jerseys, pullovers, slipovers, cardigans, etc. knitted or
crocheted (8451) was the most exported good in both 2000 and 2006. Interestingly, the
difference in export value between the most exported good and the others is a lot larger than in
the US case. The exports to the EU market thus seem less diversified.


Table 5.2 - Most exported clothing categories to the EU in USD million
                                                                       1994      2000      2006
8451      Jerseys, pullovers, slip-overs, cardigans, etc. knitted or    22.57    107.07    161.96
          crocheted
          of wool or fine animal hair                                    87%        73%       n.a.
          of cotton                                                      10%        20%       15%
8439      Women's, girls' and infants' other outer garments of           7.84     21.58     22.58
          textile fabrics, not knitted or crocheted
          of cotton                                                      83%      88,5%       72%
          of man-made fibers                                             15%         9%       14%
8471      Clothing accessories of textile fabrics, not knitted or       23.43     16.91     21.73
          crocheted
          of cotton                                                       n.a.      n.a.      n.a.
8423      Men's and boys' trousers, breeches and the like                7.60     27.12     17.25
          of cotton                                                      97%        98%       98%
8462      Undergarments, knitted or crocheted                           10.02     24.49     15.31
          of cotton                                                     100%       100%      100%
Source: COMTRADE



Inputs for exports to the EU market also differ from the inputs used for exports to the US. Cotton
is again the most important input but wool or fine animal hair is also of great importance which
can be related to the fact that knitted products are more important for the EU market.
Unfortunately there is no available data for how much of category 8451 is produced of wool or
fine animal hair in 2006 but figures for 1994 and 2000 make one believe that this input is still

94
     Brenton and Ozden (2005) p. 8
                                                                                                     40
highly important in 2006, especially when the expansion of the category 8451 is kept in mind.
Moreover, the most exported goods to the EU do not use any synthetic fibers or other fibers
which were relatively important inputs for exports to the US. Exports to the EU consequently
seem less diversified also when it comes to the inputs used.


These first results are related to ROO. The expansion of clothing exports to the US was for
example a lot faster than the expansion to the EU. This could be a result of the more liberal
AGOA ROO that allow 3rd country fabric. The investigation of exported products to the US and
the EU also seems to suggest that liberal ROO make it easier to diversify, possibly due to the fact
that input sourcing is less restricted. The finding that goods produced for the US use more
synthetic and other fibers seems to support this theory as well.




5.3 Textile Inputs
The second step of the analysis deals with the composition and origin of inputs. Has AGOA with
its liberal ROO meant more import of textile yarn and fabrics? If so, where are the inputs
sourced from? Theory would predict that more liberal ROO would lead to imports from more
efficient low-cost producers. In this case it would mean increased imports from Asia where the
most productive textile producers are located. It is further interesting to see if inputs are sourced
from producers in the EU or US which can be a sign of a ROO effect. Lastly, input origin can
also give an indication of how successful vertical integration has been. This can answer whether
it is justifiable to use strict ROO as a tool for development policy.


Imports of textile yarn and fabrics have increased with the expansion of exports of clothing.
Figure 5.4 below shows the composition of textile inputs from 1990 to 2006. The most imported
textile good has been Textile yarn (651), followed by Cotton fabrics woven (652) and Knitted or
crocheted fabrics (655). Another observation is that Textile fabrics woven other than cotton or
man-made fibers (654) that was the most imported textile good in the middle of the 1990s, today
accounts for a relatively small percentage of total imports. AGOA seems to have benefited other
products instead. Examples of such products are Tulle lace embroidery ribbons trimmings and
other small wares (656) as well as Special textile fabrics and related products (657). In general
more different kinds of products are imported after the entry into force of AGOA than in the

                                                                                                  41
1990s. Moreover, imports of textile yarn and fabrics still show a positive trend. The growth rate
of imports is slower than in the middle of the 1990s but Madagascar has never imported as much
textile as in 2006, the last year with available data.


Figure 5.4 - Total import of textile yarn and fabrics 1990-2006
 USD million
  350
  300
  250
  200
  150
  100
   50
    0
      1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

                    659 Floor coverings etc
                    658 Made-up articles wholly or chiefly of textile materials nes
                    657 Special textile fabrics and related products
                    656 Tulle lace embroidery ribbons trimmings and other small wares
                    655 Knitted or crocheted fabrics (including tubular etc fabrics)
                                           en
                    654 Textile fabrics wov other than cotton or man-made fibres
                                   en
                    653 Fabrics wov of man-made fibres (not narrow or special fabrics)
                                          en
                    652 Cotton fabrics wov (not including narrow or special fabrics)
                    651 Textile yarn


Source: COMTRADE



The most imported product in Figure 5.4, Textile yarn (651), consists of different kinds of yarn.
The most imported kind of textile yarn since 1990 is Yarn of wool of animal hair (6512). In
2006, this sub-category accounted for 87% of textile yarn imports.95 The second most imported
yarn is Cotton yarn (6513) followed by Yarn 85% of synthetic fibers (6514). The imports of
synthetic yarn have become increasingly important since the year 2000 which coincides with the
increased exports of products of synthetic fibers to the US. For a complete picture of the
composition of category 651, see Figure A.1 in Appendix.




5.3.1 Origin of Imported Inputs
To be able to relate the use of inputs to ROO, the origin of inputs must also be examined. Figure
5.5 below shows imports of textile yarn and fabrics by origin. In the beginning of the 1990s

95
     COMTRADE and author’s calculations
                                                                                              42
   inputs were mainly sourced from the EU and Africa which can be expected from the cumulation
   rules of the Lomé/Cotonou ROO and the fact that exports to the EU took off in the beginning of
   the 1990s. Imports from the US are virtually non-existent during the whole examined period.
   AGOA has hence not benefited American textile producers. The main beneficiaries in relation to
   AGOA and input sourcing are instead Asian producers. From having been responsible for only a
   small share of total textile inputs in the early 1990s Asian producers now dominate the input
   market. Other producers have also gained slightly from the entry into force of AGOA. The 3rd
   country fabric rule consequently seems to be widely used among Malagasy clothing producers
   and one of the main explanations why exports to the US have been so successful.


   Figure 5.5 - Total import of textile yarn and fabrics 1990-2006 by origin
   USD million

   350

   300

   250
                                                                                                                                       others
   200                                                                                                                                 U.S.
                                                                                                                                       EU
   150                                                                                                                                 Asia
                                                                                                                                       Africa
   100

    50

        0
            1990

                   1991

                          1992

                                 1993

                                         1994

                                                1995

                                                       1996

                                                              1997

                                                                      1998

                                                                             1999

                                                                                    2000

                                                                                            2001

                                                                                                    2002

                                                                                                           2003

                                                                                                                  2004

                                                                                                                         2005

                                                                                                                                2006




Source: COMTRADE


   Figure 5.6 - Textile yarn and fabric imports from Africa 1990-2006
   USD million
   70
   60
   50
   40
   30
   20
   10
   0
        1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006


                                        651     652     653          654     655      656          657     658      659
  Source: COMTRADE

                                                                                                                                                43
Figure 5.7 - Textile yarn and fabric imports from the EU 1990-2006
USD million
  60

  50

  40

  30

  20

  10

   0
       1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

                         651      652    653    654   655   656    657    658    659

Source: COMTRADE


Figure 5.8 - Textile yarn and fabric imports from Asia 1990-2006
USD million
 250

 200

 150

 100

  50

   0
       1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

                            651    652    653   654   655   656   657    658   659
  Source: COMTRADE



Looking more closely at the types of textile inputs sourced from Africa, the EU and Asia, certain
patterns can be revealed. Figures 5.6, 5.7 and 5.8 show textile yarn and fabric imports from
Africa, the EU and Asia respectively broken down at a 3 digit level. Note the different scales of
the figures. The y axis of Figure 5.8 for Asia goes up to USD 250 million while the y axis of
Figures 5.6 and 5.7 only reach USD 70 million and USD 60 million respectively. Also, see
Figure 5.4 if more explanation of textile categories 651-659 is wanted.


Imports of African inputs took off in the early 1990s and show a positive trend until the crisis
year of 2002, as can be seen in Figure 5.6. This result can be related to the Lomé/Cotonou ROO



                                                                                              44
that demand inputs to be sourced from the EU, domestically or another African ACP-country.96
Input sourcing from Africa would hence be allowed for products destined for the EU market.
African inputs were also most important in the 1990s during the expansion of EU exports. As
clothing exports to the EU later stagnated in the late 1990s, so did imports from Africa. As
mentioned, the crisis year of 2002 marks a shift when it comes to African inputs. First of all,
imports from Africa have never completely recovered from the crisis. Second, the composition
of imports changed. African textile inputs were dominated by three products until 2001: Textile
yarn (651), Cotton fabrics woven (652) and Knitted or crocheted fabrics (655). Since 2002,
imports have become more diversified and in particular category Fabrics woven of man-made
fibers (653) has become more important. That imports from Africa never recovered from the
crisis can be due to the fact that several Mauritian companies left Madagascar during 2002 and
never returned. The political instability was considered to be too damaging for business.97 The
increased diversification of imports can have been affected by the leaving of some Mauritian
companies but it could also be an effect of AGOA. As already has been established, exports
designated for the US use more and more of other fibers than cotton and wool.


Even if input imports from Africa have experienced a decrease in importance since 2002, the
strict Lomé/Cotonou ROO seem to have benefited African LDC producers and increased South-
South trade which would be positive from a development perspective. A closer look reveals that
Mauritius, that has outsourced much of its clothing production to Madagascar, accounts for at
least 98% of imports from Africa from 1990 to 2003. From 2004 to 2006, Mauritius’ share is
down to about 94% because of an increase of textile imports from South Africa.98 This further
highlights the importance of the leaving of Mauritian companies in 2002. The leading position of
Mauritius puts Madagascar’s African input sourcing in a slightly different light but the increased
input trade between Madagascar and Mauritius is nonetheless an example of South-South trade
that has been positively affected by the Lomé/Cotonou ROO.


Imports of EU inputs show a somewhat similar pattern to the import of African inputs which
again can be related to the strict cumulation rules of the Lomé/Cotonou ROO and to the fact that
clothing exports to the EU started in the beginning of the 1990s. Close historical ties between

96
   De Melo and Portugal-Perez (2008) p. 4
97
   Cling, Razafindrakoto and Roubaud (2004) p.20
98
   COMTRADE
                                                                                               45
Madagascar and Europe and the early French investments in the Malagasy T&C sector can also
explain the input imports from the EU. Imports of inputs from the EU showed signs of stagnation
as early as in the mid-1990s and import volumes of European textile have actually decreased by
about 20% since 1998. This is in turn a reflection of the relative stagnation of clothing exports to
the EU market. Moreover, the entry into force of AGOA has not benefited European textile
producers. This was an expected result since imports from the EU are relatively expensive so if
you can source inputs from anywhere in the world, it is rather unlikely that EU inputs would be
chosen.


Textile imports from the EU have a different composition to the African or the Asian inputs. By
far the most imported textile good from the EU since 1993 has been Textile fabrics woven other
than cotton or man-made fibers (654), a product that is basically not imported from Africa at all.
That fabric and not yarn is the most imported good from the EU is not that surprising when the
strict ROO for non-knitted products are kept in mind. Further, there is no substantial change in
the composition of imports since the middle of the 1990s. This is not surprising since the
composition of exports to the EU also has remained about the same.


Finally, as Figure 5.8 shows, textile imports from Asia did not really take off until the end of the
1990s which as mentioned above can be related to the entry into force of AGOA. That many
producers focused on the American market are financed by Asian investors should also have
simplified the sourcing of inputs from Asia.99 In comparison to imports from Africa and the EU,
imports from Asia recovered very quickly from the crisis in 2002 which can be related to the fast
recovery of exports to the US. Imports from Asia are also still increasing even if exports to the
US have decreased after 2004. The composition of imports is to some extent similar to that of
imports from Africa, but Textile yarn (651) is a more important good among imports from Asia
than among those from Africa. Tulle lace embroidery ribbons trimmings and other small wares
(656) are on the other hand a less important category than for imports from both Africa and the
EU.


The Asian inputs are foremost sourced from China but India, Singapore and Pakistan are other
countries that have contributed to the increase of Asian inputs in Malagasy clothing

99
     Cling, Razafindrakoto and Roubaud (2005) p.787
                                                                                                 46
production.100 Input sourcing from Asia is dominated by a single country to almost the same
degree as input sourcing from Africa was dominated by Mauritius but the volumes of inputs
sourced from Asia are much larger than volumes from Africa. AGOA ROO must therefore be
considered to be more favorable for South-South trade than Lomé/Cotonou ROO and above all
less trade distorting since Chinese textile producers are known to be the most efficient at the
moment.101




5.3.2 Domestic Inputs
To asses the success of vertical integration in the Malagasy T&C sector, an examination of
domestic input supplies is needed in addition to the above analysis of imported inputs. The
previous section showed that imports of textile fabrics and yarn have increased substantially
which seems to suggest that the degree of vertical integration in Madagascar has been low. If this
is the case, using strict ROO as a tool for development policy has not been a successful strategy.


Over 80% of the raw materials used in garment production are today imported. There are simply
not enough domestic supplies available in terms of both quantity and quality.102 Certain fabrics
used in clothing production do not have any domestic producers at all. The main input that can
be sourced domestically is cotton but the share of domestic cotton used in clothing production
has decreased since the middle of the 1990s. In 1993 only 34% of the Malagasy cotton demand
was imported. As the garment sector grew, cotton fabric production did not, due to the closure of
several textile mills. This was in turn an effect of the falling cotton production which above all
depended on falling cotton prices but also on harder growing conditions.103 The lowest level of
international cotton prices coincided with the crisis year of 2002 which put many cotton
producers out of business.104 The Malagasy cotton industry is today more or less destroyed and it
is hard to see that it will recover in the near future. This is considered to be a great loss for the
country since many jobs could have been created in the countryside.105



100
    COMTRADE
101
    Cadot and Nasir (2001) p. 7
102
    Global Development Solutions (2007) p. 57
103
    The region of Tulear where most cotton is grown is fighting both soil erosion and the lowest precipitation levels
in Madagascar. This in combination with a poor irrigation system makes it hard to grow anything at all.
104
    Global Development Solutions (2007) pp. 8-9
105
    Interview Delegation of the European Commission in Madagascar
                                                                                                                   47
This confirms the suspicions of a low degree of vertical integration. Not only has it not been
successful, the situation has even gotten worse. In conclusion, the Lomé/Cotonou ROO have
hardly contributed to an increase of usage of domestic inputs. There has consequently not been
any form of infant industry protection through the ROO either. This result supports the critique
of using ROO as a tool for development policy. Creating vertical integration in a competitive
sector such as T&C where low-cost inputs are essential to stay competitive is hence very difficult
in a developing country. A representative for a private sector organization further stated that
forced vertical integration does not work in Madagascar; the country is too small and the global
T&C industry is too fragmentized. Development of the cotton industry must thus be demand
driven which means that first you create a cut and sew industry and when this is working
properly people interested in making a profit will come and invest in spinning and weaving on
their own.106



5.4 Utilization Rates
Another way to investigate the impact of preferential trading arrangements and the impact of
their ROO is to look at utilization rates. Utilizations rates show exports under a preferential
trading arrangement as shares of total export to the country in question. Studies have previously
shown that a high value of preferences offered increases the probability that preferences are
requested. It has been found that the difference between preferential and 3rd country tariff rates
must be at least 4% in order for traders to have incentives for requesting preferences. If the
difference between tariff rates is lower than that, the costs of obtaining preferences are expected
to be higher than the gains from getting preferential access. The majority of the costs are found
in cost of documentation and administration of ROO. ACP countries like Madagascar are
expected to face higher costs due to among other things information disadvantages and
institutional difficulties.107


AGOA has, as seen above, had a large impact on the Malagasy T&C sector. The success
of AGOA has been related to liberal ROO. This leads to an assumption of high utilization
rates for AGOA apparel. Table 5.4 confirms this assumption by showing that the utilization rates
for AGOA apparel are very high. The high utilization rates prove that the AGOA regulations are

106
      Interview GEFP
107
      Manchin (2005) pp. ii, 21
                                                                                                48
not too costly to comply with for clothing producers. The liberal ROO have probably contributed
to this result. Low utilization rates would on the other hand mean that the procedures for getting
preferential access are too complex and costly which in turn would make producers export as 3rd
country producers instead. Another connection to ROO is that the 3rd country fabric rule for
AGOA apparel is heavily used, as much as 99% of all AGOA apparel uses 3rd country fabric.
The liberal ROO that allow 3rd country fabric for LDCs have consequently been a prerequisite
for the AGOA success.


Table 5.3 - Utilization rates for AGOA and Lomé/Cotonou apparel 2002-2007

                                                 2002        2003           2004   2005   2006   2007
Utilization rate for AGOA apparel                 84%         95%           97%    99%    96%    97%
Proportion using 3rd country fabric of            91%         92%           95%    99%    99%    99%
AGOA apparel
Utilization rate for Lomé/Cotonou apparel         85%         82%           82%    81%    81%    88%
Source: agoa.info, Eurostat and author’s calculations


Utilization rates for Lomé/Cotonou apparel are overall lower than the utilization rates for
AGOA. The only exception is 2002 but this crisis year, as has already been mentioned, cannot be
seen as representative for clothing exports in general. Other years, utilization rates for
Lomé/Cotonou are more than 10 percentage points lower than those for AGOA. Input
restrictions in the Lomé/Cotonou ROO make exporting to the EU more costly than exporting to
the US. Producers consequently choose to export as 3rd country producers to a larger extent for
exports going to the EU market than for exports going to the US market. It should also be kept in
mind that the MFN tariff for clothing in the EU ranges between 6.5%-12%. The most imported
items, Jerseys, pullovers, cardigans, waistcoats and similar articles, knitted or crocheted face
for example a 12% import tariff.108 The minimum 4% difference between preferential and 3rd
country import tariff rates is therefore met since apparel under Lomé/Cotonou enters the EU duty
free. That still almost 20% of Malagasy apparel exports to the EU were exported without using
the possibility of preferential access from 2003 to 2006 is a sign of the high costs of the
Lomé/Cotonou regulations and then in particular the high costs of the ROO.




108
      http://exporthelp.europa.eu/thdapp/taric/TaricServlet?languageId=EN
                                                                                                   49
 6      Conclusions


Textile and clothing often offers an opportunity for developing countries to diversify into
manufactured products. Madagascar is one of several African countries that have lately become
successful clothing exporters to the EU and the US. Part of the success is due to preferential
access to the EU and US markets. The preferential trading arrangements of the EU,
Lomé/Cotonou, and the US, AGOA, differ however when it comes to ROO for clothing. AGOA
has very liberal rules for input-sourcing and allows Madagascar to source inputs from anywhere
in the world (3rd country fabric rule). The ROO of Lomé/Cotonou are on the other hand stricter
and demand that inputs must be sourced domestically, from Africa or from the EU. This study
has tried to analyze the impact of these different ROO on the Malagasy textile and clothing
industry by looking at export and import patterns as well as utilization rates.


Several conclusions are drawn from the study. First, clothing exports to the EU tend to be less
diversified and contain more knitted products than exports to the US. Second, products destined
for the US market are made of more diversified inputs than those destined for the EU market.
The inputs used for exports to the EU are for example mainly cotton and wool while products for
the US market, in addition to cotton and wool, also are made of man made and synthetic fibers.
Third, exports to the US have grown more rapidly than exports to the EU after the entering into
force of AGOA with its liberal ROO in 2000. Fourth, imports of textile fabric and yarn from
Asia have increased dramatically since the year 2000. When the Malagasy T&C sector mainly
exported to the EU, i.e. before 2000, inputs were only sourced from Asia to a limited extent. The
majority of inputs were instead sourced from Africa or the EU. This is consistent with the
theory’s predictions. The ROO of Lomé/Cotonou limited Madagascar’s input sourcing
possibilities which led to supply-switching towards European and African producers due to strict
cumulation rules. When the more liberal ROO of AGOA were introduced inputs were instead
sourced from the most efficient producer on the world market, Asia. The possibility to source
inputs from Asia to competitive prices made it easier to expand and to get materials that
otherwise would have been too expensive or simply not available.


                                                                                              50
A few additional conclusions were also drawn. AGOA’s ROO were found to promote South-
South trade to a larger extent and to be less trade distorting than the Lomé/Cotonou ROO.
Economic inefficiencies are avoided through AGOA since inputs are allowed to be sourced from
the cheapest and most efficient producer on the world market. Moreover, utilization rates for
AGOA have been higher than the utilization rates for Lomé/Cotonou indicating that the ROO in
Lomé/Cotonou are perceived as complicated and costly. The 3rd country fabric rule is also
heavily used for products exported under AGOA which even further underlines the importance
of flexible input sourcing rules for expansion possibilities. Finally, no signs of increased vertical
integration were found. The usage of domestic inputs has instead decreased substantially and the
domestic cotton industry has almost been destroyed.           To use strict ROO as a tool for
development policy in the T&C sector hence does not seem to have been effective in
Madagascar.


The future of the Malagasy T&C sector is uncertain. Recent events in Madagascar and
international trade can highly affect the sector’s future, both positively and negatively. The
liberal ROO in the IEPA have created new incentives for European investments and are expected
to increase exports to the EU. One manager stated for example that her company already had
gotten new clients thanks to the new ROO and that exporting to the EU now was much easier.109
It is still too early to see any statistical effects of the IEPA but it is a possible future research
topic. Further, South Africa has thanks to SADC become more and more interested in Malagasy
clothing products. Exports to South Africa are therefore also expected to increase.


There are nonetheless obstacles that can hinder a prosperous future of the T&C sector. Current
weaknesses are for example the high cost of energy, the long distance to the world markets and
the poor domestic infrastructure. The main problem is however political instability. The current
Malagasy political crisis has had a high price for the T&C sector. Firms have gone out of
business due to lost contracts and companies have even been burnt down during riots.110 As if
this was not enough, the major threat to the industry right now is that Madagascar risks losing its
AGOA eligibility in January 2010. Since AGOA eligibility requirements include the rule of law
and political pluralism, the recent Malagasy coup d’état is not accepted by the Obama


109
      Interview MIW
110
      Interview Text’île Mada
                                                                                                  51
administration. If free elections are not held before the end of the year AGOA eligibility will
therefore most probably be lost. That would mean the end of Malagasy clothing exports to the
US and possibly the beginning of the end of the entire T&C sector. Up to 100 000 jobs are
expected to be lost indirectly if AGOA eligibility is lost and American clients are not expected to
return because of the expiration of AGOA’s 3rd country fabric rule in 2012.111


The possibility to draw any general conclusions from the results of a case study is limited. What
nonetheless can be established is that, in the case of Madagascar, ROO have had the power to
affect both clothing exports and textile input imports. The 3rd country fabric rule of AGOA has
led to a considerable expansion of the Malagasy T&C sector and diversification both in terms of
final goods produced and inputs used.       Consequently, liberal ROO have helped to create
thousands of needed formal jobs in Madagascar and thereby promoted development.




111
      Interview GEFP
                                                                                                52
References
AfDB/OECD (2005) African Economic Outlook 2004/2005, African Development Bank and the
Organisation for Economic Co-operation and Development, Tunis and Paris.


Ahmad, Munir (2007) Impact of Origin Rules for Textiles and Clothing on Developing
Countries, ICTSD Programme on Competitiveness and Sustainable Development, International
Centre for Trade and Sustainable Development, Issue Paper no. 3, Geneva.


Augier, Patricia; Gasiorek, Michael and Lai-Tong, Charles (2004) “Rules of Origin and the EU-
Med Partnership: The Case of Textiles”, The World Economy, 27(9): 1449-1473.


Augier, Patricia; Gasiorek, Michael and Lai-Tong, Charles (2005) "The Impact of Rules of
Origin on Trade Flows", Economic Policy, 20: 567-624.


Brenton, Paul and Manchin, Miriam (2003) “Making EU Trade Agreements Work: The Role of
Rules of Origin on Trade Flows", The World Economy, 26: 755-769.


Brenton, Paul and Ozden, Caglar (2005) Trade Preferences for Apparel and the Role of Rules of
Origin – The Case of Africa, The World Bank, Washington D.C.


Brenton, Paul (2006) “Enhancing Trade Preferences for LDCs: Reducing the Restrictiveness of
Rules of Origin” In: Newfarmer, Richard (eds.) Trade, Doha, and Development: A Window into
the Issues, p 277-283, The World Bank, Trade Department, Washington D.C.


Cadot, Olivier and Nasir, John (2001) Incentives and Obstacles to Growth: Lessons from
Manufacturing Case Studies in Madagascar, Regional Program on Enterprise Development
Discussion Papers, RPED Paper no. 117, The World Bank, Washington D.C.


Cadot, Olivier; Djiofack and De Melo, Jaime (2008) “Préférences commerciales et règles
d’origine : perspectives des Accords de Partenariat Économique pour l’Afrique de l’Ouest et
centrale”, Revue d’économie de développement, 2008/3, 22 :5-48.
                                                                                           53
Carrère, Céline and de Melo, Jaime (2004) Are Different Rules of Origin Equally Costly?
Estimates from NAFTA, CEPR Discussion Paper No. 4437, Centre for Economic Policy
Research, London.


Cling, Jean-Pierre; Razafindrakoto, Mireille and Roubaud, François (2004) Export Processing
Zones in Madagascar: an Endangered Success Story, Document de Travail DT/2004/02,
Développment Institutions & Analyses de Long terme (DIAL), Paris.


Cling, Jean-Pierre; Razafindrakoto, Mireille and Roubaud, François (2005) “Export Processing
Zones in Madagascar: a Success Story under Threat?”, World Development, vol. 33, no. 5, pp.
785-803.


Cling, Jean-Pierre; Razafindrakoto, Mireille and Roubaud, François (2007) Export Processing
Zones in Madagascar: the impact of the dismantling of clothing quotas on employment and
labour standards, Document de Travail DT/2007-06, Développment Institutions & Analyses de
Long terme (DIAL), Paris.


De Melo, Jaime and Portugal-Perez, Alberto (2008) Rules of Origin, Preferences and
Diversification in Apparel: African Exports to the US and to the EU, Discussion Paper Series no.
7072, Centre for Economic Policy Research, London.


Estevadeordal, Antoni and Suominen, Kati. (2003) Rules of Origin: A World Map, PEEC trade
forum, Washington D.C.


Falvey, Rod and Reed, Geoff (1998) “Economic Effects of Rules of Origin”, Review of World
Economics, 134(2), 209-229.


Ghoneim, Ahmed Farouk (2003) “Rules of Origin and Trade Diversion: The Case of the
Egyptian-European Partnership Agreement”, Journal of World Trade, 37(3): 597–622.




                                                                                             54
Global Development Solutions LLC (2007) Integrated Value Chain Analysis for the Cotton-to-
Garment Sector in Madagascar, Prepared for the Government of Madagascar and the World
Bank.


Krishna, Kala (2005), Understanding Rules of Origin, NBER Working Paper No. 11150,
Cambridge.


Krueger, Anne (1993) Free Trade Agreements as Protectionist Devices: Rules of Origin, NBER
Working Paper No. 4352, Cambridge.


Limantour, Jean-François (2006) Le Secteur Textile-Habillement de Madagascar et son
Environnement International, Cabinet JEXCO, 7 ACP 753-RPR Project n° 23a, L’Unité de
Gestion du Programme d’Appui à l’Intégration des Etats ACP dans le Système Commercial
Multilatéral, la Commission Européenne, Brussels.


Manchin, Miriam (2005) Preference utilization and tariff reduction in European Union imports
from Africa, Caribbean, and Pacific countries, Policy Research Working Paper 3688, The World
Bank, Washington D.C.


Naumann, Eckart (2008) Rules of Origin and EPAs: What has been agreed? What does it mean?
What next?, Trade Law Center for Southern Africa (Tralac), Stellenbosch


Nicita, Alessandro (2006) Export led growth, pro poor or not? Evidence from Madagascar’s
textile and apparel industry, Policy Research Working Paper 3841, The World Bank, Washington
D.C.


Nordås, Hildegunn Kyvik (2004) The Global Textile and Clothing Industry post the Agreement on
Textiles and Clothing, Discussion paper no. 5, World Trade Organization, Geneva.


Portugal-Perez, Alberto (2008) The costs of rules of origin in apparel: African preferential
exports to the United States and the European Union, Policy Issues in International Trade and



                                                                                         55
Commodities Study Series No. 39, United Nations Conference on Trade and Development, New
York and Geneva.


PricewaterhouseCoopers (2006) Sustainability Impact Assessment (SIA) of the EU-ACP Economic
Partnership Agreements – Phase Three. Rules of Origin in the Southern African Development
Community Group (SADC Group), written with financial assistance from The European
Commission, Neuilly-sur-Seine.


Rydberg, Paul Jr (2009) “AGOA Apparel Imports Resume Downward Trend in February 2009”,
US – Africa Trade Report, 12(4):1-9.


Watson, James (2007) “East Africa : a splintered picture”, Trade Negotiation Insights, Vol. 6,
No. 8, December 2007 and January 2008, International Centre for Trade and Sustainable
Development and European Centre for Development Policy Management.


Electronic Sources
AGOA info managed by TRALAC: Trade Law Centre for Southern Africa
>http://www.agoa.info/index.php?view=about&story=about
2009-01-13

>http://www.agoa.info/index.php?view=about&story=country_eligibility
2009-01-14

>http://www.agoa.info/index.php?view=about&story=apparel_rules
2009-01-14

>http://www.agoa.info/index.php?view=trade_stats&story=apparel_quotas
2009-01-14

>http://www.agoa.info/index.php?view=trade_stats&story=apparel_trade
2009-01-14


ECDPM: European Centre for Development Policy Management
>http://www.acp-eu-trade.org/index.php?loc=epa/background.php
2009-01-13


EU: European Union
>http://ec.europa.eu/development/geographical/cotonou/lomegen/lomeitoiv_en.cfm
                                                                                           56
2009-01-13

>http://ec.europa.eu/trade/issues/bilateral/regions/acp/nepa_en.htm
2009-01-13

>http://trade.ec.europa.eu/doclib/docs/2008/march/tradoc_138457.pdf
2009-01-13

>http://ec.europa.eu/development/geographical/cotonouintro_en.cfm?CFID=2311138&CFTOK
EN=de5549ec566e53bc-44BE1EAC-BCAD-6AE3- 85FE869240E498A7&jsessionid =243062
fb88384a375d62
2009-01-13

>http://trade.ec.europa.eu/doclib/docs/2007/november/tradoc_136959.pdf
2009-02-04

>http://exporthelp.europa.eu/thdapp/taric/TaricServlet?languageId=EN
2009-02-20


GEFP: Groupement des Entreprises Franches et Partenaires
>http://www.gefp.com/statistiques.php?id=secteurs
2009-01-07

>http://www.gefp.com/statistiques.php?id=nationalites_proprietaires
2009-01-07


MEFB: Ministère des Finance et du Budget Madagascar
>http://www.mefb.gov.mg/don_stat/don_sec/t12_nbzfis.pdf
2009-01-07


World Bank
>http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/AFRICAEXT/MADAGASCA
REXTN/0,,menuPK:356362~pagePK:141132~piPK:141107~theSitePK:356352,00.html
2008-10-07


UNCTAD: United Nations Conference on Trade and Development
>http://www.unctad.org/Templates/Page.asp?intItemID=2309&lang=1
2008-09-15


UNDP: United Nations Development Program
>http://hdr.undp.org/en/statistics/
2008-10-07

                                                                                   57
Interviews
- Mr. Hargreaves, John: GEFP. Vice President                            2009-05-11
- Mr. Lubineau, Vincent: ARAWAK. Manager                                2009-05-06
- Ms. Rajaonarison, Joelle: Delegation of the European Commission in
  Madagascar. In charge of trade and economy issues at the Delegation   2009-05-14
- Mr. Raveloson, Mick: Text’île Mada. General Secretary                 2009-05-05
- Ms. Vassel, Nasreen: MIW (Madagascar Industrial Washing). Manager     2009-05-12


Data sources
Agoa.info
http://www.agoa.info/index.php?view=trade_stats&story=apparel_trade

COMTRADE
http://comtrade.un.org/

Eurostat
ec.europa.eu/eurostat

OTEXA
http://otexa.ita.doc.gov/




                                                                                     58
Appendix

Table A.1 – Clothing exports of African and Asian producers in million USD after the MFA phase-out

                   2003       2004       2005       2006    2006/2004
                                                            (%)
Mauritius
Total               896        865         712        721            -16,6
EU                  627        638         555        602
USA                 269        227         167        119

Madagascar
Total               347        520         506        526             1,2
EU                  151        197         229        288
USA                 196        323         277        238

Lesotho
Total               394        457         392        388            -15,1
EU                    1          1           1          1
USA                 393        456         391        387

Kenya
Total               192        280         278        265             -5,4
EU                    4          3           7          1
USA                 188        277         271        264

China
Total            13 970     18 288     45 365      49 981        173,3
EU               12 361     13 730     22 960      22 974
USA              11 609     14 558     22 405      27 067

Vietnam
Total             3 115      3 478       3 793      4 611              33
EU                  631        758         912      1 215
USA               2 484      2 720       2 881      3 396

Cambodia
Total              1 726      2 085      2 319      2 841              36
EU                   475        643        592        690
USA                1 251      1 442      1 727      2 151
Source: Adopted from Cling, Razafindrakoto and Roubaud (2007) p. 8




                                                                                                     59
Figure A.1 – Composition of Malagasy import of textile yarn 1990-2006
Mn USD

  140

  120

  100

   80

   60

   40

   20

   0
         1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

                     6519 Yarn of textile fibres, nes
                     6518 Yarn of regenerated fibres, put up for retail sale
                     6517 Yarn of regenerated fibres, not for retail, monofil, strip, etc
                     6516 Yarn containing less than 85% of discontinuous synthetic fibres
                     6515 Yarn containing 85% or more of synthetic fibres, put up for retail
                     6514 Yarn 85% of synthetic fibres, not for retail; monofil, strip, etc
                     6513 Cotton yarn
                     6512 Yarn of wool or animal hair (including wool tops)
                     6511 Silk yarn and spun from noil or waste; silkworm gut


Source: COMTRADE




                                                                                                60
MINOR FIELD STUDY SERIES (A complete list can be ordered from Department
of Economics, University of Lund, Box 7082, S-220 07 Lund, Sweden)


170 Jonas Widmark         Nature-Based Tourism in the Cuyabeno Reserve of
                          Ecuador – A Choice Experiment Survey. 2006:2

171 Annika Bergman        FDI and spillover effects in the Indian pharmaceutical
                          industry. 2006:3

172 Cecilia Eriksson      Convergence in GDP Growth Rates across the
    Annika Persson        Provinces of China. 2006:4

173 Martin Dahlberg       Spillover Effects of FDI in the Service Sector in
                          Morocco – A Case Study of the Emerging Call-Centre
                          Industry. 2006:5

174 Christian Svensson    Dynamics of Special Economic Areas in South China –
                          Industrial Upgrading and Improving Working
                          Conditions. 2007:1

175 Josefina Danielsson   Explaining the Success of the Indian IT Industry.
    Johanna Persson       2007:2

176 Maria Holmlund        Income Poverty – A Case Study of Sri Lanka
                          2007:3

177 Emmylou Tuvhag        The Costa Rican Experience of Fair Trade Coffee –
                          Impact on producers and producer organisations
                          2007:4

178 Mikael Fridell        Exploring the Roles of Informal, Formal and
                          Semiformal Microcredit in Jordan. 2007:5

179 Åsa Hindenborg        On the Role of Remittances in Microfinance –
                          Creating Transnational Financial Services in El
                          Salvador. 2007:6

180 Linna Palmqvist       Corporate Social Responsibility and Development – A
                          case study of the CSR strategies of international
                          companies in India. 2007:7
181 Peter Holmvall       Remittances and Poverty – A case study of the
                         Philippines. 2007:8

182 Charlotta Undén      Multinational Corporations and Spillovers in Vietnam –
                         Adding Corporate Social Responsibility. 2007:9

183 Emil Samnegård       How Does Fairtrade Affect the Market? A Case Study
                         in South Africa. 2007:10

184 Emma Svensson        Microfinance, Financial Systems and Economic
                         Growth – A Theoretical Framework and Findings from
                         Bolivia. 2007:11

185 Sonja Luoma          A Path to a Life of Dreams? A Study on Financial
    Maria Persson        Markets, Microcredit and Gender in Uganda. 2008:1

186 Gabriel Andréasson   Evaluating the Effects of Economic Sanctions against
                         Burma. 2008:2

187 Katarina Öjteg       Socio-economic Determinants of HIV in Zambia – A
                         District-level Analysis. 2009:1

188 Tora Bäckman         Fairtrade Coffee and Development – A Field Study in
                         Ethiopia. 2009:2

189 Sara Forssell        Rice Price Policy in Thailand – Policy Making and
                         Recent Developments. 2009:3

190 Tora Hammar          Trade Facilitation in Vietnam – Recent Progress and
                         Impact. 2009:4

191 Joakim Persson       The Impact of a Quota System on Women’s
                         Empowerment – A Field Study in West Bengal, India.
                         2009:5

192 Erudita Hoti         Remittances and Poverty in Albania. 2009:6


193 Anna Andersson       Made in Madagascar – The Impact of Rules of Origin
                         on the Textile and Clothing Industry. 2009:7

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:5
posted:9/13/2010
language:English
pages:63